To see the other types of publications on this topic, follow the link: Two - three and four layers models.
Journal articles on the topic 'Two - three and four layers models'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Two - three and four layers models.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Rowe, Avery. "Effect of drainage layers on water retention of potting media in containers." PLOS ONE 20, no. 2 (2025): e0318716. https://doi.org/10.1371/journal.pone.0318716.
Full textAbstract:
Excess water retention in the potting medium can be a significant problem for plants grown in containers due to the volume of saturated medium which forms above the drainage hole. Adding a layer of coarse material like gravel or sand at the bottom is a common practise among gardeners with the aim of improving drainage, but some researchers have argued that such layers will raise the saturated area and in fact increase water retention. Two different depths and four different materials of drainage layer were tested with three different potting media to determine the water retention in the container after saturating and draining freely. For loamless organic media, almost all types of drainage layer reduced overall water retention in the container compared to controls. For loam-based media, most drainage layers had no effect on the overall water retention. Two simple models were also used to estimate the water retention in the media alone, excluding the drainage layer itself. All drainage layers reduced water retention of loamless organic media, according to both models. There was disagreement between the two models applied to loam-based media, and further study is required to determine the most accurate. Both models showed that some drainage layers with smaller particle sizes reduced water retention in loam-based media, but disagreed on the effect of drainage layers with larger particle sizes. Overall, any drainage layer was likely to reduce water retention of any medium, and almost never increased it. Thicker drainage layers were more effective than thinner layers, with the most effective substrate depending on the potting media used. A 60 mm layer of coarse sand was the most universally-effective drainage layer with all potting media tested.
2
Alhassan, Seiba, Gaddafi Abdul-Salaam, Michael Asante, Yaw Missah, and Ernest Ganaa. "Analyzing Autoencoder-Based Intrusion Detection System Performance." Journal of Information Security and Cybercrimes Research 6, no. 2 (2023): 105–15. http://dx.doi.org/10.26735/ylxb6430.
Full textAbstract:
The rise in cyberattacks targeting critical network infrastructure has spurred an increased emphasis on the development of robust cybersecurity measures. In this context, there is a growing exploration of effective Intrusion Detection Systems (IDS) that leverage Machine Learning (ML) and Deep Learning (DL), with a particular emphasis on autoencoders. Recognizing the pressing need to mitigate cyber threats, our study underscores the crucial importance of advancing these methodologies. Our study aims to identify the optimal architecture for an Intrusion Detection System (IDS) based on autoencoders, with a specific focus on configuring the number of hidden layers. To achieve this objective, we designed four distinct sub-models, each featuring a different number of hidden layers: Test 1 (one hidden layer), Test 2 (two hidden layers), Test 3 (three hidden layers), and Test 4 (four hidden layers).We subjected our models to rigorous training and testing, maintaining consistent neuron counts of 30 and 60. The outcomes of our experimental study reveal that the model with a single hidden layer consistently outperformed its counterparts, achieving an accuracy of 95.11% for NSL-KDD and an impressive 98.6% for CIC-IDS2017. The findings of our study indicate that our proposed system is viable for implementation on critical network infrastructure as a proactive measure against cyber-attacks.
3
Qdroo, Awf, and Muhammet Baykara. "A New Approach to Detect Fake News Related to Covid-19 Pandemic Using Deep Neural Network." Journal of Applied Science and Technology Trends 3, no. 02 (2022): 20–27. http://dx.doi.org/10.38094/jastt302124.
Full textAbstract:
The fake news that accompanied the COVID-19 pandemic on social media platforms negatively affected people and led to a state of panic and fear of the unknown. This study aims to build a model for classifying textual news for four datasets related to COVID-19, binary classification (fake and real) with high performance. Two hybrid deep learning models were built. The first model consists of three layers of a one-dimension convolutional neural network (1D-CNN), followed by two layers of a long-short-term memory neural network (LSTM). The second model consists of three layers of a 1D-CNN followed by two layers of a bidirectional LSTM neural network (BiLSTM). Finally, the results obtained using hybrid models were compared with the results obtained by applying three machine learning classifiers (naïve Bayes, logistic regression, and k-nearest neighbor) on the same data sets. This study achieved promising results with an accuracy of (96.98%, 94.52%, 99.60%, and 99.90%) for the first model with all data sets and (97.15%, 95.32%, 99.40%, and 99.82%) for the second model with the same four data sets.
4
Schubert, Wayne, Richard Taft, and Christopher Slocum. "A Simple Family of Tropical Cyclone Models." Meteorology 2, no. 2 (2023): 149–70. http://dx.doi.org/10.3390/meteorology2020011.
Full textAbstract:
This review discusses a simple family of models capable of simulating tropical cyclone life cycles, including intensification, the formation of the axisymmetric version of boundary layer shocks, and the development of an eyewall. Four models are discussed, all of which are axisymmetric, f-plane, three-layer models. All four models have the same parameterizations of convective mass flux and air–sea interaction, but differ in their formulations of the radial and tangential equations of motion, i.e., they have different dry dynamical cores. The most complete model is the primitive equation (PE) model, which uses the unapproximated momentum equations for each of the three layers. The simplest is the gradient balanced (GB) model, which replaces the three radial momentum equations with gradient balance relations and replaces the boundary layer tangential wind equation with a diagnostic equation that is essentially a high Rossby number version of the local Ekman balance. Numerical integrations of the boundary layer equations confirm that the PE model can produce boundary layer shocks, while the GB model cannot. To better understand these differences in GB and PE dynamics, we also consider two hybrid balanced models (HB1 and HB2), which differ from GB only in their treatment of the boundary layer momentum equations. Because their boundary layer dynamics is more accurate than GB, both HB1 and HB2 can produce results more similar to the PE model, if they are solved in an appropriate manner.
5
Tsepav, Matthew Tersoo, Azeh Yakubu, Kumar Niranjan, et al. "Geophysical Characterisation of Native Clay Deposits in Some Parts of Niger State, Nigeria." Journal of Physics: Theories and Applications 6, no. 1 (2022): 43. http://dx.doi.org/10.20961/jphystheor-appl.v6i1.56457.
Full textAbstract:
<p class="Abstract">Clay minerals are among the world’s most important and useful industrial minerals. Conductance, transmissivity and corrosity are some physical parameters for determining quality clay. Four (4) clay deposit sites in Kaffin-Koro, Dutse, Dogon-Ruwa and Kushikoko were investigated to evaluate corrosivity, the longitudinal conductance and transmissivity to determine the clay quality. Electrical resistivity method employing Schlumberger electrode array was used to determine the thicknesses and depths of the subsurface strata while Interpex 1xD software was used to interpret the data. Three (3) to four (4) layer earth models were delineated. Kaffin-Koro and Dutse showed three layer models while Dogon-Ruwa and Kushikoko revealed four layers. Moderate clay content was found in Kaffin-Koro in the second layer with longitudinal conductance value of 0.4780 siemens and thickness 0.770m at depth of 1.17m Dogon-Ruwa also had moderate clay content in the third layer with conductance value of 0.237 siemens, depth of 2.43m and thickness 1.76m. Kushikoko had low clay deposit in the second layer with conductance 0.1810 siemens and thickness 2.73 m at 4.37 m while the clay deposit in Dutse appeared to be generally poor as the longitudinal conductance values of the top two layers were less than 0.1 siemens.</p>
6
Deokar, A. H., and A.A. Korake. "Experimental Investigation of Ferrocement Slab Panels Using Square Metal Mesh." Journal of Structural Engineering, its Applications and Analysis 7, no. 3 (2024): 7–12. https://doi.org/10.5281/zenodo.13770899.
Full textAbstract:
<em>The ongoing assessment depicts the outcomes of testing level ferrocement sheets upheld with different quantities of wire network layer and assortment in board thickness. The essential objective of these exploratory tests is to consider the effect of using different amounts of wire network Layers and thickness minor takeoff from the flexural strength of stage ferrocement discussions and to concentrate on the impact of fluctuating the amount of twine local area layers on the malleability and a conclusive force of this type of Ferrocement structure. In this assessment, all of the models were parceled into four social events to explore the strength and lead of ferrocement level sheets presented to two-point stacking. 24 ferrocement added substances have been built and endeavored. 24 ferrocement components were built and tried. The pre-owned number of wire network layers is single, two, three and four layers; additionally, thicknesses of boards are 30mm and 40mm.</em>
7
Hekal, Ghada Mousa, Ayman Magdy Moawad Elshaboury, and Yousry B. I. Shaheen. "The impact of openings on ferrocement I-beams: a study on metallic and non-metallic mesh reinforcement." Challenge Journal of Concrete Research Letters 15, no. 2 (2024): 30. http://dx.doi.org/10.20528/cjcrl.2024.02.001.
Full textAbstract:
The primary objective of this investigation is to assess the influence of openings on the structural performance of ferrocement I-beams, incorporating diverse metallic and non-metallic mesh reinforcements. Sixteen beams underwent testing utilizing a four-point loading system until failure, categorized into four groups based on the type of mesh reinforcement. Each group comprised a control I-beam without openings and three additional beams featuring one, two, and three openings, respectively. To ensure consistent reinforcement weight, the four groups were reinforced with three layers of welded steel meshes, two layers of expanded metal meshes, two layers of Tensar meshes, and eight layers of Gavazzi meshes. Comparative analysis of the experimental outcomes was conducted with finite element models utilizing Abaqus. Therefore, there was good agreement between the experimental and numerical results. The findings showed that beams with no openings, one, and two openings reinforced with Gavazzi meshes had the highest ultimate load compared to other tested beams, while beams with three openings, those reinforced with expanded metal meshes had the greatest ultimate loads. Placing three openings in beams, with dimensions of 100×50 mm (two of these openings are approximately 10 cm apart from each edge while the third opening is located at mid-span), reduced the load-to-weight ratio by about 20.7%, 12.9%, 8.2%, and 23.8% for welded beams, expanded beams, Tensar beams, and Gavazzi beams, respectively, compared to the beams with no openings.
8
Baines, Peter G., and Fiona Guest. "The nature of upstream blocking in uniformly stratified flow over long obstacles." Journal of Fluid Mechanics 188 (March 1988): 23–45. http://dx.doi.org/10.1017/s002211208800062x.
Full textAbstract:
The general method described in Baines 1988 has been applied to stratified flows of finite depth over long obstacles where the flow initially has uniform horizontal velocity. The fluid consists of a finite number of homogeneous layers of equal thickness and with equal density increments. This represents the state of continuous stratification with constant density gradient as closely as possible, for a given number of layers. Two-, three-, four- and sixty-four-layered models are studied in detail. The results are expressed in terms of the initial Froude number F0 (F0 = U/ĉ1 where U is the fluid speed and ĉ1 is the speed of the fastest long internal wave mode in the fluid at rest) and the obstacle height. In general, introduction of an obstacle into the flow causes disturbances to propagate upstream (columnar disturbance modes) which alter the velocity and density profiles there. These may accumulate to cause upstream blocking of some of the fluid layers if F0 is sufficiently small. As the number of fluid layers increases, so does the range of F0 for which this upstream blocked flow occurs. There are no upstream disturbances for F0 > 1, and for F0 < 1 the upstream disturbances are of the rarefaction type if upstream blocking does not occur. The results for three and four layers show how several coexisting modes may interact to affect the upstream profiles. The results for sixty-four-layers provide theoretical support for the observational criterion (Baines 1979b) that blocking in initially uniformly stratified flow occurs when Nhm/U > 2 (N is the Brunt—Väisälä frequency and hm the obstacle height), provided that more than two modes are present. In some situations, layered models are found to be inadequate as a representation of continuous stratification when one or more layers thicken to the extent that their discreteness is significant.
9
Lawrence, E., E. J. Garba, Y. M. Malgwi, and M. A. Hambali. "An Application of Artificial Neural Network for Wind Speeds and Directions Forecasts in Airports." European Journal of Electrical Engineering and Computer Science 6, no. 1 (2022): 53–59. http://dx.doi.org/10.24018/ejece.2022.6.1.407.
Full textAbstract:
Wind speed patterns are highly dynamic and non-linear and thus cannot be accurately forecasted using conventional linear regression models. In this work, Artificial Neural Network (ANN) technique was applied to forecast wind speeds and directions in airports. Monthly data of maximum temperature, minimum temperature, wind speed, wind direction, relative humidity and wind run for Yola International Airport were collected from 1995 to 2021 from Nigerian Meteorological Agency (NIMET) Abuja-Nigeria. Six Neural Network models were built. ANN with no hidden layers, ANN model with one hidden layer and two dropout layers, ANN model with four hidden layers and three dropout layers, ANN model with eight hidden layers, ANN model with nine hidden layers and finally, ANN model with ten hidden layers. Back Propagation training algorithm was implemented using the PYTHON toolbox. Each of the models was trained using the training dataset and validated using the validation dataset. To test the forecasting ability of each of the models we tested it using unknown data that is the test dataset. The results from each of the models were organized and assessed in terms of the magnitude of the statistical error between the measured result and the real data. This was achieved by measuring the average of the Mean Square Errors (MSE) and Mean Absolute Error (MAE) for each of the models used for forecasting both wind speeds and directions. The results show that Multilayer perceptron with ten hidden layers with (MSE) = 0.92 and (MAE) = 0.73 emerged as the most preferred model for wind speeds forecast while the multilayer perceptron with four hidden layers with (MSE) = 1,858 and (MAE) = 35 emerged the most preferred model for wind directions forecast. Future research can be carried out to improve the accuracy of the model for wind direction forecasts.
10
Mehmood, Maryam, Farhan Hussain, Ahsan Shahzad, and Nouman Ali. "Classification of Remote Sensing Datasets with Different Deep Learning Architectures." Earth Sciences Research Journal 28, no. 4 (2025): 409–19. https://doi.org/10.15446/esrj.v28n4.113518.
Full textAbstract:
Remote sensing image classification has great advantages in the areas of environmental monitoring, urban planning, disaster management and many others. Unmanned Aerial Vehicles (UAVs) have revolutionized remote sensing by providing high-resolution imagery. In this context, effective image classification is crucial for extracting meaningful information from UAV-captured images. This study presents a comparison of different deep learning-based approach for supervised image classification of UAV images. We have experimented on four different CNN models like VGG 16, Alex net, Resnet50 and a deep neural network Efficient-Net-B0 on different remote sensing datasets; AID and AIDER. Multiple combinations were tried to find out which model performs better on which type of datasets. We have used pre-trained initial layers of four CNN models (AlexNet, VGG 16, Resnet50 and Efficient-Net-Bo) then last three layers of each of the selected models are removed and new layers have been added with better tuned parameters. Two different schemes were analyzed. In Scheme-1 the original AlexNet, VGG 16, Resnet50 and Efficient-Net-B0 were experimented without changing and tuning their number of parameters, while in Scheme-2 transfer learning was applied on the pre-trained models and after removing last three layers new layers were added with better tuned hyper-parameters. The evaluation of above schemes was ensured through comprehensive metrics across diverse land cover classes, four different performance evaluation matrices namely; F1 score, precision, accuracy and recall. The main focus of this research is towards transfer learning and adding new layers into pre-trained models to get better classification accuracy.
11
Faried, A. S., Kh Osman, and M. G. Mohamed. "The Effect of Diffrenet Types of Strengthening RC Columns Subjected to Eccentric Load." International Journal of Membrane Science and Technology 10, no. 3 (2023): 3303–15. http://dx.doi.org/10.15379/ijmst.v10i3.3281.
Full textAbstract:
In many constructions, columns are subjected to eccentric loads. So, they are unable to withstand higher loading requirements. Ferrocement jacketing and fiber-reinforced polymer (FRP) have demonstrated exceptionally high efficacy in comparison to other strengthening techniques. This paper presents the results of experimental studies on reinforced concrete columns strengthened with carbon fiber reinforced polymer (CFRP) and ferrocement layer composites under an eccentric load. The new techniques depended on the combination of (CFRP) and Ferrocement to strengthen RC columns, which are subjected to eccentric loads. To get the average result for the model, ten tested models are patterned using two of each specimen, totaling twenty tested specimens. Models are divided into four groups in accordance with the method of strengthening. The specimen in the first group has not been strengthened. The second group consists of six specimens strengthened with one, two, or three successive layers of the wire mesh with an outer 15 mm thickened coating. The third group consists of four specimens strengthened with one or two successively wrapped layers of the (CFRP). The fifth group consists of six specimens strengthened with one or two layers of CFRP and one or two layers of concrete wire mesh. The purpose of the current study is to compare the use of CFRP, ferrocement layers, or a combination of carbon fibers and ferrocement to strengthen concrete columns that are subjected to eccentric loads (e/t = 0.5). The study has also shown that mixing GFRP with ferrocement is more effective than using ferrocement or carbon, but more expensive.
12
Mustak Un Nobi, Md, Md Rifat, M. F. Mridha, Sultan Alfarhood, Mejdl Safran, and Dunren Che. "GLD-Det: Guava Leaf Disease Detection in Real-Time Using Lightweight Deep Learning Approach Based on MobileNet." Agronomy 13, no. 9 (2023): 2240. http://dx.doi.org/10.3390/agronomy13092240.
Full textAbstract:
The guava plant is widely cultivated in various regions of the Sub-Continent and Asian countries, including Bangladesh, due to its adaptability to different soil conditions and climate environments. The fruit plays a crucial role in providing food security and nutrition for the human body. However, guava plants are susceptible to various infectious leaf diseases, leading to significant crop losses. To address this issue, several heavyweight deep learning models have been developed in precision agriculture. This research proposes a transfer learning-based model named GLD-Det, which is designed to be both lightweight and robust, enabling real-time detection of guava leaf disease using two benchmark datasets. GLD-Det is a modified version of MobileNet, featuring additional components with two pooling layers such as max and global average, three batch normalisation layers, three dropout layers, ReLU as an activation function with four dense layers, and SoftMax as a classification layer with the last lighter dense layer. The proposed GLD-Det model outperforms all existing models with impressive accuracy, precision, recall, and AUC score with values of 0.98, 0.98, 0.97, and 0.99 on one dataset, and with values of 0.97, 0.97, 0.96, and 0.99 for the other dataset, respectively. Furthermore, to enhance trust and transparency, the proposed model has been explained using the Grad-CAM technique, a class-discriminative localisation approach.
13
Deng, Qi. "Blockchain Economical Models, Delegated Proof of Economic Value and Delegated Adaptive Byzantine Fault Tolerance and their implementation in Artificial Intelligence BlockCloud." Journal of Risk and Financial Management 12, no. 4 (2019): 177. http://dx.doi.org/10.3390/jrfm12040177.
Full textAbstract:
The Artificial Intelligence BlockCloud (AIBC) is an artificial intelligence and blockchain technology based large-scale decentralized ecosystem that allows system-wide low-cost sharing of computing and storage resources. The AIBC consists of four layers: a fundamental layer, a resource layer, an application layer, and an ecosystem layer (the latter three are the collective “upper-layers”). The AIBC layers have distinguished responsibilities and thus performance and robustness requirements. The upper layers need to follow a set of economic policies strictly and run on a deterministic and robust protocol. While the fundamental layer needs to follow a protocol with high throughput without sacrificing robustness. As such, the AIBC implements a two-consensus scheme to enforce economic policies and achieve performance and robustness: Delegated Proof of Economic Value (DPoEV) incentive consensus on the upper layers, and Delegated Adaptive Byzantine Fault Tolerance (DABFT) distributed consensus on the fundamental layer. The DPoEV uses the knowledge map algorithm to accurately assess the economic value of digital assets. The DABFT uses deep learning techniques to predict and select the most suitable BFT algorithm in order to enforce the DPoEV, as well as to achieve the best balance of performance, robustness, and security. The DPoEV-DABFT dual-consensus architecture, by design, makes the AIBC attack-proof against risks such as double-spending, short-range and 51% attacks; it has a built-in dynamic sharding feature that allows scalability and eliminates the single-shard takeover. Our contribution is four-fold: that we develop a set of innovative economic models governing the monetary, trading and supply-demand policies in the AIBC; that we establish an upper-layer DPoEV incentive consensus algorithm that implements the economic policies; that we provide a fundamental layer DABFT distributed consensus algorithm that executes the DPoEV with adaptability; and that we prove the economic models can be effectively enforced by AIBC’s DPoEV-DABFT dual-consensus architecture.
14
Chiper Titire, Larisa, and Cristian Muntenita. "Experimental and Numerical Analysis of the Damage Mechanism of an Aramid Fabric Panel Engaged in a Medium-Velocity Impact." Polymers 16, no. 13 (2024): 1920. http://dx.doi.org/10.3390/polym16131920.
Full textAbstract:
The aim of this study is to analyze the ballistic impact behavior of a panel made of Twaron CT736 fabric with a 9 mm Full Metal Jacket (FMJ) projectile. Three shots are fired at different velocities at this panel. The ballistic impact test procedure was carried out in accordance with NIJ 010106. The NIJ-010106 standard is a document that specifies the minimum performance requirements that protection systems must meet to ensure performance. The 9 mm FMJ projectile is, according to NIJ 010106, in threat level II, but the impact velocity is in threat level IIIA. Analysis of macro-photographs of the impact of the Twaron CT736 laminated fabric panel with a 9 mm FMJ projectile involves a detailed examination of the images to gather information about the material performance and failure mechanisms at the macro- or even meso-level (fabric/layer, thread). In this paper, we analyze numerically and experimentally a panel consisting of 32 layers, made of a single material, on impact with a 9 mm FMJ projectile. The experimental results show that following impact of the panel with three projectiles, with velocities between 414 m/s and 428 m/s, partial penetration occurs, with a different number of layers destroyed, i.e., 15 layers in the case of the projectile velocity of 414 m/s, 20 layers of material in the case of the panel velocity of 422 m/s and 22 layers destroyed in the case of the projectile velocity of 428 m/s. Validation of the simulated model is achieved by two important criteria: the number of broken layers and the qualitative appearance. Four numerical models were simulated, of which three models validated the impact results of the three projectiles that impacted the panel. Partial penetration occurs in all four models, breaking the panel in the impact area, with only one exception, i.e., the number of layers destroyed, in which case the simulation did not validate the validation criterion. The performance of Twaron CT736 fabric is also given by the indentation depth values by two methods: according to NIJ 0101.06 and by 3D scanning. The NIJ 010106 standard specifies that a panel provides protection when the indentation depth values are less than 0.44 mm.
15
Adedotun, Adeyemo I., Akande V. Oluwatimilehin, Akinlalu A. Adewale, and Sanusi S. Olumide. "Comparative Studies of Subsurface Layers’ Competence Evaluation using TOPSIS and AHP Models at Ilaramokin, Southwestern Nigeria." Physics Access 03, no. 02 (2023): 78–98. http://dx.doi.org/10.47514/phyaccess.2023.3.2.012.
Full textAbstract:
The incessant collapse of buildings associated with geotechnical incompetence in different parts of Nigeria and the rapid growth of Ilaramokin town near Akure Southwestern Nigeria motivated this work. Two multi-criteria decision analysis approaches were used in integrating geoelectric parameters (topsoil resistivity, weathered layer resistivity and bedrock resistivity), static water level measurements and geology in evaluating the subsurface geotechnical competence of Ilaramokin. Thirty (30) vertical electrical sounding, eighty-six (86) static water level measurements and geological maps were used. The vertical electrical sounding (VES) results delineated three to five geoelectric layers which correspond to four geologic layers namely; the topsoil, weathered layer, partially weathered/fractured basement and presumed bedrock. The resistivity of the geologic layers varies respectively from 48 - 701, 31 - 1065, 14 - 139, 132 - 6582 Ωm, while their thickness varies from 0.4 - 4.1, 1.3 - 11.6 and 4.0 - 20.1 m in the three upper layers respectively. The VES results were presented as topsoil, weathered layer and bedrock resistivity maps. The VES results, static water level measurement and geology were integrated using both the Technique for the Order of Prioritization by Similarity to Ideal Solution (TOPSIS) and Analytic Hierarchy Process (AHP) models to produce geotechnical competence maps. Consistency test and grain size analysis were carried out on 10 soil samples obtained across the area to validate the geotechnical competence model maps produced using both TOPSIS and AHP models. The validation showed that the geotechnical model map produced from the TOPSIS model has a higher percentage (90%) of correlation with consistency tests and grain size analysis compared to that of the AHP model (70%).
16
Al-Bashaireh, Khaled. "Radiocarbon Age Determinations of Mosaic Mortar Layers of Churches from North Jordan." Radiocarbon 57, no. 5 (2015): 851–63. http://dx.doi.org/10.2458/azu_rc.57.18197.
Full textAbstract:
This research is aimed at radiocarbon dating organic inclusions and lime-binder powders of mortar layers of mosaic pavements in four churches of arguable archaeological date located in northern Jordan. One mortar sample from each mosaic pavement of each church was collected, examined by thin section microscopy, and then physically pretreated by gentle crushing and dry sieving to collect lime-binder powders of different grain sizes. Charcoal samples uncovered from three samples and the CO2 gases, collected by hydrochloric acid (HCl) hydrolysis of the powders, were 14C dated using accelerator mass spectrometry (AMS). Four powders of 63–45 μm from the four samples and two powders of 45–38 μm from two samples were analyzed in order to get more precise dates and examine previous proposed models for the interpretation of the results. 14C determinations showed agreement between charcoal ages and archaeological data, while the fine lime-binder's powders, especially from the mosaic's bedding layer, produced more precise dates. Results suggest that 14C date profiles produced by HCl hydrolysis of the lime-binder powders can be clearly interpreted by the existing models.
17
Dobeš, Pavel, Antonín Lokaj, and Kristýna Vavrušová. "Stiffness and Deformation Analysis of Cross-Laminated Timber (CLT) Panels Made of Nordic Spruce Based on Experimental Testing, Analytical Calculation and Numerical Modeling." Buildings 13, no. 1 (2023): 200. http://dx.doi.org/10.3390/buildings13010200.
Full textAbstract:
Timber structures are currently more important for solving tasks in construction practice. For this reason, there is an opportunity for research in the area of physical tests and numerical models. This paper deals with the determination and comparison of the deformation properties of cross-laminated timber (CLT) panels based on laboratory tests, analytical calculation and numerical modeling. CLT panels are structural building components consisting of cross-oriented solid timber layers. Three types of panels with different geometry and number of layers (three, five and seven) were experimentally tested using a four-point bending test, where load–deformation curves were recorded. The results of the experimental testing of the three-layer panels were subsequently compared with a numerical model in SCIA Engineer, a numerical model in ANSYS Workbench and an analytical calculation. The research shows a good agreement in bending behavior between the laboratory tests, the analytical calculation according to the standard and two different approaches in numerical analysis.
18
Jafar, Raed, Adel Awad, Kamel Jafar, and Isam Shahrour. "Predicting Effluent Quality in Full-Scale Wastewater Treatment Plants Using Shallow and Deep Artificial Neural Networks." Sustainability 14, no. 23 (2022): 15598. http://dx.doi.org/10.3390/su142315598.
Full textAbstract:
This research focuses on applying artificial neural networks with nonlinear transformation (ANNs) models to predict the performance of wastewater treatment plant (WWTP) processes. The paper presents a novel machine learning (ML)-based approach for predicting effluent quality in WWTPs through explaining the relationships between the multiple influent and effluent pollution variables of an existing WWTP. We developed AI models such as feed-forward neural network (FFNN) and random forest (RF) as well as deep learning methods such as convolutional neural network (CNN), recurrent neural network (RNN), and pre-train stacked auto-encoder (SAE) in order to avoid various shortcomings of conventional mechanistic models. The developed models focus on providing an adaptive, functional, and alternative methodology for modeling the performance of the WWTP. They are based on pollution data collected over three years. It includes chemical oxygen demand (COD), biochemical oxygen demand (BOD5), phosphates (PO₄−3), and nitrates (NO₃−), as well as auxiliary indicators including the temperature (T), degree of acidity or alkalinity (pH), electric conductivity (EC), and the total dissolved solids (TDS). The paper presents the results of using SNN- and DNN-based models to predict the effluent concentrations. Our results show that SNN can predict plant performance with a correlation coefficient (R) up to 88%, 90%, 93%, and 96% for the single models COD, BOD5, NO₃−, and PO₄−3, respectively, and up to 88%, 96%, and 93% for the ensemble models (BOD5 and COD), (PO₄−3 and NO₃−), and (COD, BOD5, NO₃−, PO₄−3), respectively. The results also show that the two-hidden-layers model outperforms the one-hidden-layer model (SNN). Moreover, increasing the input parameters improves the performance of models with one and two hidden layers. We applied DNN (CNN, RNN, SAE) with three, four, and five hidden layers for WWTP modeling, but due to the small datasets, it gave a low performance and accuracy. In sum, this paper shows that SNN (one and two hidden layers) and the random forest (RF) machine learning technique provide effective modeling of the WWTP process and could be used in the WWTP management.
19
Nuñez Ponasso, Guillermo, William A. Wartman, Ryan C. McSweeney, et al. "Improving EEG Forward Modeling Using High-Resolution Five-Layer BEM-FMM Head Models: Effect on Source Reconstruction Accuracy." Bioengineering 11, no. 11 (2024): 1071. http://dx.doi.org/10.3390/bioengineering11111071.
Full textAbstract:
Electroencephalographic (EEG) source localization is a fundamental tool for clinical diagnoses and brain-computer interfaces. We investigate the impact of model complexity on reconstruction accuracy by comparing the widely used three-layer boundary element method (BEM) as an inverse method against a five-layer BEM accelerated by the fast multipole method (BEM-FMM) and coupled with adaptive mesh refinement (AMR) as forward solver. Modern BEM-FMM with AMR can solve high-resolution multi-tissue models efficiently and accurately. We generated noiseless 256-channel EEG data from 15 subjects in the Connectome Young Adult dataset, using four anatomically relevant dipole positions, three conductivity sets, and two head segmentations; we mapped localization errors across the entire grey matter from 4000 dipole positions. The average location error among our four selected dipoles is ∼5mm (±2mm) with an orientation error of ∼12∘ (±7∘). The average source localization error across the entire grey matter is ∼9mm (±4mm), with a tendency for smaller errors on the occipital lobe. Our findings indicate that while three-layer models are robust under noiseless conditions, substantial localization errors (10–20mm) are common. Therefore, models of five or more layers may be needed for accurate source reconstruction in critical applications involving noisy EEG data.
20
Yun, Jong-Hwan, Gun-Woong Yoon, Yu-Jae Jeon, and Min-Soo Kang. "Evaluation of the Properties of 3D-Printed Onyx–Fiberglass Composites." Materials 17, no. 16 (2024): 4140. http://dx.doi.org/10.3390/ma17164140.
Full textAbstract:
This study evaluated the properties of 3D-printed Onyx–fiberglass composites. These composites were 3D-printed with zero, one, two, three, and four layers of fiberglass. Ten samples of each configuration were printed for the tensile and flexural tests. The average tensile strength of the Onyx specimens was calculated to be 44.79 MPa, which increased linearly by approximately 20–25 MPa with each additional fiberglass layer. The elastic moduli calculated from the micromechanics models were compared with the experimental values obtained from the tensile tests. The experimental elastic modulus increased more significantly than the model prediction when more fiberglass layers were added. The flexural modulus of Onyx was 17.6 GPa, which increased with each additional fiberglass layer. This quantitative analysis of composites fabricated using 3D printing highlights their potential for commercialization and industrial applications.
21
Ruckley, C. V., J. J. Dale, B. Gibson, D. Brown, A. J. Lee, and R. J. Prescott. "Multi-layer compression: comparison of four different four-layer bandage systems applied to the leg." Phlebology: The Journal of Venous Disease 18, no. 3 (2003): 123–29. http://dx.doi.org/10.1258/026835503322381324.
Full textAbstract:
Objective: To compare on standardized laboratory models the performance of four commercially available four-layer bandage systems. Methods: Four experienced bandagers applied each of the four systems [Profore® Regular (Smith & Nephew, Hull, UK), Ultra Four (Robinsons, Chesterfield, UK), System 4 (SSL International, Knutsford, UK) and K-Four® (Parema, Loughborough, UK)] to two models: a 12.5 cm diameter padded cylinder and a 9.5-14.5 cm padded cone. Bandages were applied individually in single layers and as a completed system using standard application techniques. Pressures were measured by the Borgnis Medical Stocking Tester at positions corresponding to ankle, gaiter and mid-calf areas as determined by the pressure sensor. Results: A total of 768 observations were made: 384 for each model, 192 for each bandaging system, 192 for each bandager and 128 for each measuring point. The increase in pressure produced by each additional layer was in the range of 50-60% of the pressure achieved by the same bandage when used as a single layer. Each bandage system and each bandager produced a gradient of final mean pressure irrespective of whether the bandage was applied to a cylinder or a cone. However, there were no significant differences in the gradients between the four bandage systems or between the four bandagers. There were significant differences in the final pressures achieved among the bandage systems when applied as completed systems (mean: Profore® = 42 mmHg; System 4 = 45 mmHg; K-Four® = 48 mmHg; and Ultra Four = 51 mmHg; P<0.001). Conclusions: These results challenge a commonly-held assumption concerning the additive effect of pressures generated by successive bandage layers. When applied as part of a multi-layered system each bandage adds just over half the pressure achieved by the same bandage when applied alone. The four completed systems produced pressures within a range appropriate for ulcer therapy, although there were significant differences in mean pressures. This capability of the systems to produce different pressures could be clinically important in the hands of inexperienced bandagers or with patients at risk of pressure damage..
22
Pérez-Castillo, José Luis, Angel Mora, Rogelio Perez-Santiago, Armando Roman-Flores, Rafiq Ahmad, and Enrique Cuan-Urquizo. "Flexural Properties of Lattices Fabricated with Planar and Curved Layered Fused Filament Fabrication." Materials 16, no. 9 (2023): 3451. http://dx.doi.org/10.3390/ma16093451.
Full textAbstract:
The use of curved layers in fused filament fabrication could lead to various advantages in surface finishing and mechanical properties. Here, the influence of three different structural and manufacturing parameters (volume fraction, raster arrangement, and the use of curved or planar layers) on the mechanical properties of lattice structures under three-point bending is studied. Two different raster arrangements were considered, i.e., those with rasters at planes parallel to the principal axes of the samples and those diagonally arranged, all at four different volume fractions. All different samples were additively manufactured using planar and curved layers. Samples were further dimensionally inspected to refine the computational models before their analysis via finite element simulations. The linear elastic region of the load-displacement curves was further analyzed numerically via finite element models. Predictions with finite element models resulted in good agreement with errors below 10%. Samples with diagonal rasters were 70% softer than those parallel to the principal axes.
23
Gómez Franco, Maribel, Antonio Ramírez Treviño, Eduard Figueras, and Angel Sauceda Carvajal. "Thermoelectrical characterization and comparative analysis of three finite element models of a MEMS thermal sensor." Superficies y Vacío 31, no. 2 (2018): 33–38. http://dx.doi.org/10.47566/2018_syv31_1-020033.
Full textAbstract:
This document presents the finite element modeling using ANSYS to obtain the thermal resistance of a MEMS thermal sensor. Additionally, the document describes a thermoelectrical characterization to find the sensor performance parameters. For modeling purposes, we divided the thermal sensor into four different thickness zones. We analyzed three different models, the first includes all materials layers, the second involves an equivalent thermal conductivity and an equivalent thickness for each zone, and the proposed model besides using an equivalent thermal conductivity by zone also considers the same thickness for all zones to reduce simulation time and to optimize thermal sensor design parameters. The first model evaluates three different boundary conditions, while the second and third models consider two different thermopile wide strips. The thermal resistance of the proposed model has a relative error of 11% in relation to the experimental value. The model, considering all layers and heat power applied to the surface as boundary conditions, has the lowest error (9%), while models considering the thermopile strips width have shown a higher error, 67%. As a result, the proposed model for heat transfer analysis simplifies complex geometries and reduces simulation time.
24
Abubakar, Aminuddeen, and Ahmed Baita Garko. "A Predictive Model for Network Intrusion Detection System Using Deep Neural Network." Dutse Journal of Pure and Applied Sciences 7, no. 3a (2022): 113–28. http://dx.doi.org/10.4314/dujopas.v7i3a.12.
Full textAbstract:
Network Intrusion Detection System (NIDS) is an important part of Cyber safety and security. It plays a key role in all networked ICT systems in detecting rampant attacks such as Denial of Service (DoS) and ransom ware attacks. Existing methods are inadequate in terms of accuracy detection of attacks. However, the requirement for high accuracy detection of attacks using Deep Neural Network requires expensive computing resources which in turn make most organisations, and individuals shy away from it. This study therefore aims at designing a predictive model for network intrusion detection using deep neural networks with very limited computing resources. The study adopted Cross Industry Standard Process for Data Mining (CRISP-DM) as one of the formal methodologies and python was used for both testing and training, using crucial parameters such as the learning rate, number of epochs, neurons and hidden layers which greatly determined the accuracy level of the DNN algorithm. These parameters were experimented with values that are lesser compared to previous studies, training and evaluation were also done on the KDD99 data-set. The varying values of accuracy obtained from this study on four models with different numbers of layers of 50-epochs and learning rate of 0.01 achieved competitive results in comparison with the previous research of 100-1000 epochs and learning rate of 0.1. Therefore, the model with two layers attained same accuracy of 0.955 as the model with three layers from the previous study out of the four models tested in this study.
 Also, the models with three and four layers in this study attained an accuracy of 0.956, which is 0.001greater than the previous study's models.
 Keywords: Network-Based IDS, Host-Based IDS, Deep Neural Network, Denial of Service, Knowledge Discovery Dataset
25
Aradea, Rianto, and Husni Mubarok. "Recognition of Organic Waste Objects Based on Vision Systems Using Attention Convolutional Neural Networks Models." Scientific Journal of Informatics 11, no. 3 (2024): 595–608. http://dx.doi.org/10.15294/sji.v11i3.6494.
Full textAbstract:
Purpose: High population growth and increasing consumption patterns have resulted in significant organic waste production. The public often does not understand the correct way to deal with the problem of organic waste, including public awareness regarding the need for its management. Therefore, a system is needed to recognize waste objects based on various types. Currently, much research in this field has been studying object recognition, for example, the implementation of the Convolutional Neural Networks (CNN) model. However, there are still various challenges that must be addressed, including objects with diverse visual characteristics such as form, size, color, and physical condition. This research focuses on developing a system that enhances object recognition of waste, specifically organic waste, using an Attention Convolutional Neural Network (ACNN). By integrating attention mechanisms into the CNN model, this study addresses the challenges of recognizing waste objects with diverse visual characteristics. The proposed system seeks to improve the accuracy and efficiency of organic waste identification, which is crucial for advancing waste management practices and reducing environmental impact. Methods: This research combines a CNN architecture with an attention mechanism to create a better object detection environment called Attention-CNN (ACNN). The ACNN architecture employed consists of one layer input, three convoluted layers, three max-pooling layers, one attention layer, one flattened layer, four dropout layers, and two dense layers arranged in a certain way. Result: The research result shows that the model CNN with attention mechanism (ACNN) was slightly better at 86.93% than the standard model of CNN, which accounted for 86.70% in accuracy. Novelty: In general, the current use of CNN architecture to address waste object recognition problems typically employs standard architectures, resulting in lower accuracy for complex waste objects. In contrast, our research integrates attention mechanisms into the CNN architecture (ACNN), enhancing the model's ability to focus on relevant features of waste objects. This leads to improved recognition accuracy and robustness against visual variability. This distinction is important as it overcomes the limitations of standard CNN models in handling visually diverse and complex waste objects, thereby highlighting the novelty and contribution of our research.
26
Yang, Yang, Siqin Liu, Fangbo Chen, et al. "Exploring multiple electrical layers overlying coal seams using the transient electromagnetic method." PLOS ONE 17, no. 10 (2022): e0273423. http://dx.doi.org/10.1371/journal.pone.0273423.
Full textAbstract:
The transient electromagnetic method (TEM) is widely applied in coal hydrogeological exploration owing to its sensitivity to low-resistivity bodies. However, when a coal seam is buried deep, particularly if there are multiple electrical layers in the vertical direction of the overlying stratum, the results of the calculation using the late-channel empirical formula of the TEM may no longer reflect the actual situation. In this study, we evaluated the principle of the one-dimensional (1D) Occam algorithm and the steps for performing an inversion. We proposed various two-, three-, and four-layer electrical models for inversion using the 1D Occam algorithm. Our results are consistent with the electrical distribution of the models, thus indicating the effectiveness of the algorithm. A test project of large-depth transient electromagnetic exploration in the Datong Coalfield in Shanxi, China, was selected for experimental verification. The 1D Occam inversion was used to successfully identify various electrical strata overlying the coal seam.
27
Alouffi, Bader, Abdullah Alharbi, Radhya Sahal, and Hager Saleh. "An Optimized Hybrid Deep Learning Model to Detect COVID-19 Misleading Information." Computational Intelligence and Neuroscience 2021 (November 15, 2021): 1–15. http://dx.doi.org/10.1155/2021/9615034.
Full textAbstract:
Fake news is challenging to detect due to mixing accurate and inaccurate information from reliable and unreliable sources. Social media is a data source that is not trustworthy all the time, especially in the COVID-19 outbreak. During the COVID-19 epidemic, fake news is widely spread. The best way to deal with this is early detection. Accordingly, in this work, we have proposed a hybrid deep learning model that uses convolutional neural network (CNN) and long short-term memory (LSTM) to detect COVID-19 fake news. The proposed model consists of some layers: an embedding layer, a convolutional layer, a pooling layer, an LSTM layer, a flatten layer, a dense layer, and an output layer. For experimental results, three COVID-19 fake news datasets are used to evaluate six machine learning models, two deep learning models, and our proposed model. The machine learning models are DT, KNN, LR, RF, SVM, and NB, while the deep learning models are CNN and LSTM. Also, four matrices are used to validate the results: accuracy, precision, recall, and F1-measure. The conducted experiments show that the proposed model outperforms the six machine learning models and the two deep learning models. Consequently, the proposed system is capable of detecting the fake news of COVID-19 significantly.
28
van Till, Sietske A. L., Mariia V. Maksimova, Ghislaine J. M. W. van Thiel, and Eline M. Bunnik. "An assessment of the moral value of neuronal cell models and brain organoids." Molecular Psychology: Brain, Behavior, and Society 2 (July 10, 2023): 15. http://dx.doi.org/10.12688/molpsychol.17557.1.
Full textAbstract:
Advances in stem cell technology enable neuroscientists to develop induced pluripotent stem cell (iPSC)-based neuronal models of varying complexity, ranging from single human brain cells to two-dimensional neuronal cell models and three-dimensional brain organoids. While the discussion on the moral status of brain organoids is taking center stage in the bioethical literature and is invariably linked to the presumed capacity of future brain organoids to develop some form of consciousness, analyses of the moral status of other – less complex – iPSC-based neuronal models are lacking. In this paper we aim to clarify the moral value of various types of existing neuronal models, including brain organoids. We show how it is made up of several layers that may encompass various sorts of considerations, including moral values, the results of empirical research, and biological characteristics. We identify four such layers – instrumental, intrinsic, symbolic, and relational – that are relevant for the assessment of the moral value of neuronal models. We demonstrate that it lies not in a capacity to develop some form of consciousness (which is absent in current iPSC-based neuronal models, including brain organoids), but in other considerations, including the genetic links between models and donors, the ability of models to mimic brain (dys)function, and their symbolic value, all of which are often overlooked in the bioethical literature. Also, we demonstrate that the 'thickness' of the layers (i.e., their moral weight) increases when the neuronal model is more complex. Finally, we discuss the practical-ethical implications of our analysis for the use of neuronal models in research settings, for instance in relation to informed consent and biobank governance. Our four-layer framework can be applied also in moral assessments of other iPSC-based models, including emerging and future cell models.
29
Lin, Yaoqiang, Mengfei Zhou, Nan Wang, et al. "Optimization of the Target Layer for Three-Dimensional Shale Gas Development in Weiyuan Block." E3S Web of Conferences 598 (2024): 01008. http://dx.doi.org/10.1051/e3sconf/202459801008.
Full textAbstract:
In order to explore whether the Weiyuan block can effectively carry out three-dimensional development well network deployment, we should first explore whether there is a target layer to meet the three-dimensional development. Based on three-dimensional geological, fracture and rock mechanics models, this paper finds that the absolute stress difference between Longyi14a minor layer and Longyi14b minor layer in the study area is less than 5 MPa, and there is no obvious stress isolation section, and the longitudinal distance of the upper part of Longyi14a minor layer from the top of Longyi11 layer distributes in the range of 18~28 m, which is able to effectively avoid strong inter-well interference with the horizontal wells of Longyi11 layer. So, according to the distribution characteristics of the remaining reserves and reservoir characteristics of Weiyuan block, Longyi14a minor layer is initially preferred as the target for three-dimensional development. In order to further explore the specific layers, based on the original three wells of H3 platform, four threedimensional development schemes with different target layers were designed, deploying L4 and L5 wells in the Longyi13 layer~Longyi14c minor layer, and the Longyi14a minor layer was confirmed as a three-dimensional development target layer system based on numerical simulation results. Finally, in the lower half branch of H3 platform, an “M” type three-dimensional well network pattern with five wells was adopted (three 4a wells of the lower layer Longyi11 layer L1, L2 and L3, and two wells of the upper Longyi14a minor layer - L4 and L5). The three-dimensional well network can increase production for the platform by 1.76×108 m3 and recovery increased from 14.3% to 29.3%, an overall increase of 15%. The relevant understanding can provide a reference for the three-dimensional development strategy of Weiyuan shale gas block.
30
Mazurkiewicz, Łukasz, Jerzy Małachowski, Krzysztof Damaziak, Paweł Baranowski, and Paweł Gotowicki. "IDENTIFICATION OF LAYERS DISTRIBUTION IN THE COMPOSITE COUPON USING FINITE ELEMENT METHOD AND THREE POINT BENDING TEST." Acta Mechanica et Automatica 7, no. 3 (2013): 160–65. http://dx.doi.org/10.2478/ama-2013-0027.
Full textAbstract:
Abstract The main objective of the study is to develop experimentally validated FE model and perform numerical analysis of layered composites made by hand lay-up techniques during tension and bending test. The research object is glass - polyester laminate made of four unidirectional layers. In order to validate the numerical models experimental test were performed. Due to the very different stiffness modulus in tension and bending loading the material properties obtained from standard test are not suitable to apply in numerical model. Significantly different behaviour compared to experimental test was obtained for tree point bending where the numerical model becomes too stiff. Simple coupons, relatively easy to manufacture presented in the paper have very low quality. The differences in actual and theoretical bending stiffness (obtained from tension stiffness) exceed 70%. In order to represent the actual structure the layers of the composite were divided by resin layers and also additional resin layer at the top and bottom of the model were defined. Single stage optimization process was used to adjust the material layout. After layer set-up modification very significant improvement can be seen for flexural behaviour
31
Jahantigh, Moslem, and Hamidreza Ramazi. "KNOWLEDGE DRIVEN METHODS FOR CU-AU PORPHYRY POTENTIAL MODELLING; A CASE STUDY OF THE MOKHTARAN AREA, EASTERN IRAN." Rudarsko-geološko-naftni zbornik 39, no. 3 (2024): 131–44. http://dx.doi.org/10.17794/rgn.2024.3.10.
Full textAbstract:
Current research investigates multi-criteria decision methods, consisting of AHP TOPSIS, AHP VIKOR and AHP MOORA, to model porphyry copper potential in the Mokhtaran area in Eastern Iran. Evidential layers in this study include intrusive rocks, volcanic rocks, faults, Geochemical mineralization probability index (GMPI), reduction to the magnetic pole of the total magnetic intensity map, argillic and phyllic alterations. The importance of these evidential layers was calculated using the AHP method. Then, a fuzzy method was applied to the same scale the evidential layers. The threshold values of these layers were discretized with the Fractal method. Then, a weight was assigned to each evidential layer. After weighing all of the evidential layers, different MCDM methods, including AHP TOPSIS, AHP VIKOR, and AHP MOORA, were implemented to combine these layers and outline the Porphyry Copper Prospectivity Models. The predicted models show the same promising areas. The appropriate coincidence can be seen between high potential areas and mine indications. Then the success curve rate was implemented to compare the three predicted models. Based on this method, the AHP TOPSIS has a better performance. Since the success rate curve belongs to AHP TOPSIS, it is placed above the other two methods. Next, AHP VIKOR has a better performance than AHP MOORA. The three MCDM methods produced the same Cu porphyry mineralization areasd along fault zones.
32
Al-Shourbaji, Ibrahim, Pramod H. Kachare, Laith Abualigah, et al. "A Deep Batch Normalized Convolution Approach for Improving COVID-19 Detection from Chest X-ray Images." Pathogens 12, no. 1 (2022): 17. http://dx.doi.org/10.3390/pathogens12010017.
Full textAbstract:
Pre-trained machine learning models have recently been widely used to detect COVID-19 automatically from X-ray images. Although these models can selectively retrain their layers for the desired task, the output remains biased due to the massive number of pre-trained weights and parameters. This paper proposes a novel batch normalized convolutional neural network (BNCNN) model to identify COVID-19 cases from chest X-ray images in binary and multi-class frameworks with a dual aim to extract salient features that improve model performance over pre-trained image analysis networks while reducing computational complexity. The BNCNN model has three phases: Data pre-processing to normalize and resize X-ray images, Feature extraction to generate feature maps, and Classification to predict labels based on the feature maps. Feature extraction uses four repetitions of a block comprising a convolution layer to learn suitable kernel weights for the features map, a batch normalization layer to solve the internal covariance shift of feature maps, and a max-pooling layer to find the highest-level patterns by increasing the convolution span. The classifier section uses two repetitions of a block comprising a dense layer to learn complex feature maps, a batch normalization layer to standardize internal feature maps, and a dropout layer to avoid overfitting while aiding the model generalization. Comparative analysis shows that when applied to an open-access dataset, the proposed BNCNN model performs better than four other comparative pre-trained models for three-way and two-way class datasets. Moreover, the BNCNN requires fewer parameters than the pre-trained models, suggesting better deployment suitability on low-resource devices.
33
Devi, N. "Offline Handwritten Character Recognition using Convolutional Neural Network." International Journal for Research in Applied Science and Engineering Technology 9, no. 8 (2021): 1483–89. http://dx.doi.org/10.22214/ijraset.2021.37610.
Full textAbstract:
Abstract: This paper focuses on the task of recognizing handwritten Hindi characters using a Convolutional Neural Network (CNN) based. The recognized characters can then be stored digitally in the computer or used for other purposes. The dataset used is obtained from the UC Irvine Machine Learning Repository which contains 92,000 images divided into training (80%) and test set (20%). It contains different forms of handwritten Devanagari characters written by different individuals which can be used to train and test handwritten text recognizers. It contains four CNN layers followed by three fully connected layers for recognition. Grayscale handwritten character images are used as input. Filters are applied on the images to extract different features at each layer. This is done by the Convolution operation. The two other main operations involved are Pooling and Flattening. The output of the CNN layers is fed to the fully connected layers. Finally, the chance or probability score of each character is determined and the character with the highest probability score is shown as the output. A recognition accuracy of 98.94% is obtained. Similar models exist for the purpose, but the proposed model achieved a better performance and accuracy than some of the earlier models. Keywords: Devanagari characters, Convolutional Neural Networks, Image Processing
34
Gamri, K. A., та B. S. Omarov. "СРАВНИТЕЛЬНЫЙ АНАЛИЗ МЕТОДОВ ГЛУБОКОГО ОБУЧЕНИЯ ДЛЯ ВЫЯВЛЕНИЯ ПНЕВМОНИИ НА РЕНТГЕНОВСКИХ ИЗОБРАЖЕНИЯХ". INTERNATIONAL JOURNAL OF INFORMATION AND COMMUNICATION TECHNOLOGIES 3, № 4(12) (2022): 70–83. http://dx.doi.org/10.54309/ijict.2022.12.4.006.
Full textAbstract:
Pneumonia is a potentially fatal bacterial illness that affects one or both lungs in humans and is frequently caused by the bacterium Streptococcus pneumoniae. According to the World Health Organization, pneumonia accounts for one in every three fatalities in India (WHO). Expert radiotherapists must evaluate chest X-rays used to diagnose pneumonia. Thus, establishing an autonomous method for identifying pneumonia would be advantageous for treating the condition as soon as possible, especially in distant places. Convolutional Neural Networks (CNNs) have received a lot of interest for illness categorization due to the effectiveness of deep learning algorithms in evaluating medical imagery. Furthermore, features gained by pre-trained CNN models on large-scale datasets of X-ray pictures are extremely effective in image classification tasks. Several Convolutional Neural Networks were seen to categorize x-ray pictures into two groups, pneumonia and non-pneumonia, using various parameters, hyperparameters, and number of convolutional layers modified by the authors. The study analyzes six different models. The first and second models each include two and three convolutional layers. VGG16, VGG19, ResNet50 and Inception-v3 are the other four pre-trained models.
35
Kothadiya, Deep, Chintan Bhatt, Krenil Sapariya, Kevin Patel, Ana-Belén Gil-González, and Juan M. Corchado. "Deepsign: Sign Language Detection and Recognition Using Deep Learning." Electronics 11, no. 11 (2022): 1780. http://dx.doi.org/10.3390/electronics11111780.
Full textAbstract:
The predominant means of communication is speech; however, there are persons whose speaking or hearing abilities are impaired. Communication presents a significant barrier for persons with such disabilities. The use of deep learning methods can help to reduce communication barriers. This paper proposes a deep learning-based model that detects and recognizes the words from a person’s gestures. Deep learning models, namely, LSTM and GRU (feedback-based learning models), are used to recognize signs from isolated Indian Sign Language (ISL) video frames. The four different sequential combinations of LSTM and GRU (as there are two layers of LSTM and two layers of GRU) were used with our own dataset, IISL2020. The proposed model, consisting of a single layer of LSTM followed by GRU, achieves around 97% accuracy over 11 different signs. This method may help persons who are unaware of sign language to communicate with persons whose speech or hearing is impaired.
36
Hao, Yaobin, and Fangying Song. "Fourier Neural Operator Networks for Solving Reaction–Diffusion Equations." Fluids 9, no. 11 (2024): 258. http://dx.doi.org/10.3390/fluids9110258.
Full textAbstract:
In this paper, we used Fourier Neural Operator (FNO) networks to solve reaction–diffusion equations. The FNO is a novel framework designed to solve partial differential equations by learning mappings between infinite-dimensional functional spaces. We applied the FNO to the Surface Quasi-Geostrophic (SQG) equation, and we tested the model with two significantly different initial conditions: Vortex Initial Conditions and Sinusoidal Initial Conditions. Furthermore, we explored the generalization ability of the model by evaluating its performance when trained on Vortex Initial Conditions and applied to Sinusoidal Initial Conditions. Additionally, we investigated the modes (frequency parameters) used during training, analyzing their impact on the experimental results, and we determined the most suitable modes for this study. Next, we conducted experiments on the number of convolutional layers. The results showed that the performance of the models did not differ significantly when using two, three, or four layers, with the performance of two or three layers even slightly surpassing that of four layers. However, as the number of layers increased to five, the performance improved significantly. Beyond 10 layers, overfitting became evident. Based on these observations, we selected the optimal number of layers to ensure the best model performance. Given the autoregressive nature of the FNO, we also applied it to solve the Gray–Scott (GS) model, analyzing the impact of different input time steps on the performance of the model during recursive solving. The results indicated that the FNO requires sufficient information to capture the long-term evolution of the equations. However, compared to traditional methods, the FNO offers a significant advantage by requiring almost no additional computation time when predicting with new initial conditions.
37
Libersou, Sonia, Xavier Siebert, Malika Ouldali, et al. "Geometric Mismatches within the Concentric Layers of Rotavirus Particles: a Potential Regulatory Switch of Viral Particle Transcription Activity." Journal of Virology 82, no. 6 (2008): 2844–52. http://dx.doi.org/10.1128/jvi.02268-07.
Full textAbstract:
ABSTRACT Rotaviruses are prototypical double-stranded RNA viruses whose triple-layered icosahedral capsid constitutes transcriptional machinery activated by the release of the external layer. To understand the molecular basis of this activation, we studied the structural interplay between the three capsid layers by electron cryo-microscopy and digital image processing. Two viral particles and four virus-like particles containing various combinations of inner (VP2)-, middle (VP6)-, and outer (VP7)-layer proteins were studied. We observed that the absence of the VP2 layer increases the particle diameter and changes the type of quasi-equivalent icosahedral symmetry, as described by the shift in triangulation number (T) of the VP6 layer (from T = 13 to T = 19 or more). By fitting X-ray models of VP6 into each reconstruction, we determined the quasi-atomic structures of the middle layers. These models showed that the VP6 lattices, i.e., curvature and trimer contacts, are characteristic of the particle composition. The different functional states of VP6 thus appear as being characterized by trimers having similar conformations but establishing different intertrimeric contacts. Remarkably, the external protein VP7 reorients the VP6 trimers located around the fivefold axes of the icosahedral capsid, thereby shrinking the channel through which mRNA exits the transcribing rotavirus particle. We conclude that the constraints arising from the different geometries imposed by the external and internal layers of the rotavirus capsid constitute a potential switch regulating the transcription activity of the viral particles.
38
Huang, Xuxiang, Chen Xiang, Hua Li, and Peng He. "SBugLocater: Bug Localization Based on Deep Matching and Information Retrieval." Mathematical Problems in Engineering 2022 (August 24, 2022): 1–14. http://dx.doi.org/10.1155/2022/3987981.
Full textAbstract:
Bug localization is a technology that locates buggy source files using bug reports reported by users. Automatic localization of buggy files can speed up the process of bug fixing to improve the efficiency and productivity of software quality assurance teams. Nowadays, some research studies have investigated the natural language information retrieval technology, but few of them have applied the matching technology in deep learning to bug localization. Therefore, we propose a bug localization model SBugLocater based on deep matching and IR. The model composes of three layers: semantic matching layer, relevance matching layer, and IR layer. In particular, the relevance matching layer captures fine-grained local matching signals, while coarse-grained semantic similarity signals come from the semantic matching layer. Further, based on collaborative filtering in different directions, the IR layer works to find whether bug reports and source files are related, which indirectly transforms the matching task of different grammatical structures between bug reports and source files into the same structure and solves the mismatching problem of the first two matching models when the query is short. In our work, four benchmark data sets are used as experimental data sets and Accuracy@k, MAP, and MRR as evaluation metrics, which are used to compare and analyze the performance of bug localization with the four state-of-the-art methods. Experimental results show that SBugLocater outperforms the four models. For example, compared with the best of the four models, the evaluation metrics of Accuracy@10, MAP, and MRR are improved by 6.9%, 13.9%, and 17%, respectively.
39
Mitrevski, Vangelce, Cvetanka Mitrevska, Mirko Babić, Tale Geramitcioski, and Borce Mitrevski. "Convective drying kinetics of osmotically pre-treated potato slices." Journal on Processing and Energy in Agriculture 25, no. 1 (2021): 13–15. http://dx.doi.org/10.5937/jpea24-31192.
Full textAbstract:
In the food industry, convective drying is a widely used method due to its applicability to many food materials. Besides this advantage of the convective drying method, there are several shortcomings related to the rehydration capacity, low quality of dried material, loss of color, flavour, and nutrient of the final dried materials. In this paper, the convective drying kinetics of osmotically pre-treated potato slices (variety Carrera) were analyzed. Thin-layer drying kinetics of potato slices at four drying air temperatures 40, 50, 60 and 70°C and two drying air velocities 1 and 2 ms-1 were obtained on the experimental setup. For an approximation of the experimental data with regard to the moisture ratio three thin-layers drying, models from scientific literature and the model of Mitrevski et al., were used. For each model and data set the statistical performance index, ph chi-squared, and ch2, values were calculated and models were ranked afterward.
40
E.M., Keita *. B. Mbow M.L. Sow C. Sow M. Thiam C. Sene. "THEORETICAL COMPARATIVE STUDY OF INTERNAL QUANTUM EFFICIENCY OF THIN FILMS SOLAR CELLS BASED ON CuInSe2 : p+/p/n/n+, p/n/n+, p+/p/n and p/n MODELS." INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY 5, no. 9 (2016): 344–59. https://doi.org/10.5281/zenodo.154202.
Full textAbstract:
In this paper we propose to study the performances of thin films solar cells based on CuInSe<sub>2</sub>. The following models are studied : p/n ; p<sup>+</sup>/p/n ; p/n/n<sup>+</sup>; p<sup>+</sup>/p/n/n<sup>+</sup>.<sup> </sup>The objective of this work is to study the performance of the homojunction based on CuInSe<sub>2</sub>, with a medium band gap window layer based on CuInS<sub>2</sub>, deposited on a substrate, according to the model CuInS<sub>2</sub>(p<sup>+</sup>)/CuInSe<sub>2</sub>(p)/CuInSe<sub>2</sub>(n)/CuInSe<sub>2</sub>(n<sup>+</sup>) (p<sup>+</sup>/p/n/n<sup>+</sup>). We compare this structure (p<sup>+</sup>/p/n/n<sup>+</sup>) with the following models: the homojunction CuInSe<sub>2</sub>(p)/CuInSe<sub>2</sub>(n) (p/n), the homojonction with window layer CuInS<sub>2</sub>(p<sup>+</sup>)/CuInSe<sub>2</sub>(p)/CuInSe<sub>2</sub>(n) (p/n/n<sup>+</sup>) and the homojunction deposited on substrate CuInSe<sub>2</sub>(p)/CuInSe<sub>2</sub>(n)/ CuInSe<sub>2</sub>(n<sup>+</sup>) (p/n/n<sup>+</sup>). Calculation models for determining the density of the minority carriers, the density of the photocurrent, and the internal quantum efficiency were established for the different structures. These theoretical results are used to compare their performance. In order to test the validity of our calculations models, We compare our results with some experimental results published in the literature
41
Sawah, Mohamed S., Shereen Aly Taie, Mohamed Hasan Ibrahim, and Shereen A. Hussein. "An accurate traffic flow prediction using long-short term memory and gated recurrent unit networks." Bulletin of Electrical Engineering and Informatics 12, no. 3 (2023): 1806–16. http://dx.doi.org/10.11591/eei.v12i3.5080.
Full textAbstract:
Congestion on roadways is an issue in many cities, especially at peak times, which causes air and noise pollution and cause pressure on citizens. So, the implementation of intelligent transportation systems (ITSs) is a very important part of smart cities. As a result, the importance of making accurate short-term predictions of traffic flow has significantly increased in recent years. However, the current methods for predicting short-term traffic flow are incapable of effectively capturing the complex non-linearity of traffic flow that affects prediction accuracy. To overcome this problem, this study introduces two novel models. The first model uses two long-short term memory (LSTM) units that can extract the traffic flow temporal features followed by four dense layers to perform the traffic flow prediction. The second model uses two gated recurrent unit (GRU) units that can extract the traffic flow temporal features followed by three dense layers to perform the traffic flow prediction. The two proposed models give promising results on performance measurement system (PEMS), traffic and congestions (TRANCOS) dataset that is firstly used as metadata. So, the two models can do this in specific cases and are able to suddenly capture trend changes.
42
Mohamed, S. Sawah, Aly Taie Shereen, Hasan Ibrahim Mohamed, and A. Hussein Shereen. "An accurate traffic flow prediction using long-short term memory and gated recurrent unit networks." Bulletin of Electrical Engineering and Informatics 12, no. 3 (2023): 1806~1816. https://doi.org/10.11591/eei.v12i3.5080.
Full textAbstract:
Congestion on roadways is an issue in many cities, especially at peak times, which causes air and noise pollution and cause pressure on citizens. So, the implementation of intelligent transportation systems (ITSs) is a very important part of smart cities. As a result, the importance of making accurate short-term predictions of traffic flow has significantly increased in recent years. However, the current methods for predicting short-term traffic flow are incapable of effectively capturing the complex non-linearity of traffic flow that affects prediction accuracy. To overcome this problem, this study introduces two novel models. The first model uses two long-short term memory (LSTM) units that can extract the traffic flow temporal features followed by four dense layers to perform the traffic flow prediction. The second model uses two gated recurrent unit (GRU) units that can extract the traffic flow temporal features followed by three dense layers to perform the traffic flow prediction. The two proposed models give promising results on performance measurement system (PEMS), traffic and congestions (TRANCOS) dataset that is firstly used as metadata. So, the two models can do this in specific cases and are able to suddenly capture trend changes.
43
Saman, Bander, P. Gogna, El-Sayed Hasaneen, J. Chandy, E. Heller, and F. C. Jain. "Spatial Wavefunction Switched (SWS) FET SRAM Circuits and Simulation." International Journal of High Speed Electronics and Systems 26, no. 03 (2017): 1740009. http://dx.doi.org/10.1142/s0129156417400092.
Full textAbstract:
This paper presents the design and simulation of static random access memory (SRAM) using two channel spatial wavefunction switched field-effect transistor (SWS-FET), also known as a twin-drain metal oxide semiconductor field effect transistor (MOS-FET). In the SWS-FET, the channel between source and drain has two quantum well layers separated by a high band gap material between them. The gate voltage controls the charge carrier concentration in the quantum well layers and it causes the switching of charge carriers from one channel to other channel of the device. The standard SRAM circuit has six transistors (6T), two p-type MOS-FET and four n-type MOS-FET. By using the SWSFET, the size and the number of transistors are reduced and all of transistors are n-channel SWS-FET. This paper proposes two different models of the SWS-FET SRAM circuits with three transistors (3T) and four transistors (4T) also addresses the stability of the proposed SWS-FET SRAM circuits by using the N-curve analysis. The proposed models are based on integration between Berkeley Shortchannel IGFET Model (BSIM) and Analog Behavioral Model (ABM), the model is suitable to investigate the gates configuration and transient analysis at circuit level.
44
Pałczyński, Cezary, and Paweł Olejnik. "Anomalies Classification in Fan Systems Using Dual-Branch Neural Networks with Continuous Wavelet Transform Layers: An Experimental Study." Information 16, no. 2 (2025): 71. https://doi.org/10.3390/info16020071.
Full textAbstract:
In this study, anomalies in a fan system were classified using a real measurement setup to simulate mechanical anomalies such as blade detachment or debris accumulation. Data were collected under normal operating conditions and with an added unbalancing mass. Additionally, sensor anomalies were introduced by manipulating accelerometer readings and examining three types: spike, stuck, and dropout. To classify the anomalies, four neural network models—variations in Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) were tested. These models incorporated a Continuous Wavelet Transform (CWT) layer. A novel approach for implementing the CWT layer in both LSTM and CNN architectures was proposed, along with a dual-branch input structure featuring two CWT layers using different mother wavelets. The dual-branch configuration with different mother wavelets yielded better accuracy for the simpler LSTM network. Accuracy comparisons were conducted for the 10 best-performing models based on validation set predictions, revealing improved classification performance. The study concluded with a summary of prediction accuracy for both the validation and test sets of data, along with the calculation of average accuracy, demonstrating the effectiveness of the proposed dual-branch neural network structure in classifying anomalies in fan systems.
45
Makeeva, N. V. "Neural network methods for vowel classification in the vocalic systems with the [ATR] (Advanced Tongue Root) contrast." Philosophical Problems of IT & Cyberspace (PhilIT&C), no. 2 (December 18, 2023): 49–60. http://dx.doi.org/10.17726/philit.2023.2.4.
Full textAbstract:
The paper aims to discuss the results of testing a neural network which classifies the vowels of the vocalic system with the [ATR] (Advanced Tongue Root) contrast based on the data of Akebu (Kwa family). The acoustic nature of the [ATR] feature is yet understudied. The only reliable acoustic correlate of [ATR] is the magnitude of the first formant (F1) which can be also modulated by tongue height, resulting in significant overlap between high [-ATR] vowels and mid [+ATR] vowels. Other acoustic metrics which had been associated with the [ATR], such as F1 bandwidth (B1), relative intensity of F1 to F2 (A1-A2), etc., are typically inconsistent across vowel types and speakers. The values of four metrics – F1, F2, A1-A2, B1 – were used for training and testing the neural network. We tested four versions of the model differing in the presence of the fifth variable encoding the speaker and the number of hidden layers. The models which included the variable encoding the speaker achieved slightly higher accuracy, whereas the precision and recall metrics of the three-layer model were generally higher than those with two hidden layers.
46
Xie, Jupeng, Huajun Zhang, Linfan Liu, Mengchuan Li, and Yixin Su. "Decomposition-Based Multistep Sea Wind Speed Forecasting Using Stacked Gated Recurrent Unit Improved by Residual Connections." Complexity 2021 (November 15, 2021): 1–14. http://dx.doi.org/10.1155/2021/2727218.
Full textAbstract:
Sea wind speed forecast is important for meteorological navigation system to keep ships in safe areas. The high volatility and uncertainty of wind make it difficult to accurately forecast multistep wind speed. This paper proposes a new decomposition-based model to forecast hourly sea wind speeds. Because mode mixing affects the accuracy of the empirical mode decomposition- (EMD-) based models, this model uses the variational mode decomposition (VMD) to alleviate this problem. To improve the accuracy of predicting subseries with high nonlinearity, this model uses stacked gate recurrent units (GRU) networks. To alleviate the degradation effect of stacked GRU, this model modifies them by adding residual connections to the deep layers. This model decomposes the nonlinear wind speed data into four subseries with different frequencies adaptively. Each stacked GRU predictor has four layers and the residual connections are added to the last two layers. The predictors have 24 inputs and 3 outputs, and the forecast is an ensemble of five predictors’ outputs. The proposed model can predict wind speed in the next 3 hours according to the past 24 hours’ wind speed data. The experiment results on three different sea areas show that the performance of this model surpasses those of a state-of-the-art model, several benchmarks, and decomposition-based models.
47
Fayed, Sabry, Mohamed Ghalla, Ayman El-Zohairy, Ehab A. Mlybari, Rabeea W. Bazuhair, and Mohamed Emara. "Construction Efficiency in Shear Strengthening of Pre-Cracked Reinforced Concrete Beams Using Steel Mesh Reinforced Strain Hardening Cementitious Composites." Buildings 15, no. 6 (2025): 945. https://doi.org/10.3390/buildings15060945.
Full textAbstract:
Because of the degradation of building materials and the increased design load, concrete parts continually require repair. Special cementitious matrix components, Strain Hardening Cementitious Composites (SHCC), have exceptional ductility, strength growth during cracking, and recurrent controlled-opening crack formation. The purpose of this study was to improve the qualities of SHCC by reinforcing it with steel metal mesh. This study examined the optimization and effects of shear strengthening on the shear capacity of both damaged and undamaged reinforced concrete beams by employing SHCC internally reinforced with steel mesh fabric (SMF). Under bending loading, eight reinforced concrete beams were evaluated. Four of them were loaded to shear crack before any strengthening could be performed. The beams were 1500 mm in length, 200 mm in height, and 120 mm in width, and one, two, or three SMFs were applied. The beams’ whole shear span had external strengthening applied on both sides. Additionally, layers of strengthening in the U-shape were applied. The walls of the strengthening were thirty millimeters thick. The failure, load-deflection response, ultimate load, ultimate displacement, and energy absorbance of the tested beams were determined and discussed. Compared to an unstrengthened beam, the ultimate load of undamaged beams increased by 47%, 57%, and 90% when reinforced with 1, 2, or 3 layers of SMF, respectively, within the SHCC. Additionally, incorporating one, two, or three SMF layers within the SHCC improved the deflection of strengthened undamaged beams by 52%, 87%, and 116%, respectively. For damaged beams, the maximum load was approximately 11% lower than that of their undamaged counterparts, regardless of the number of SMF layers used in the SHCC strengthening. Applying one, two, or three layers of SMFs within the strengthening layer led to increases of the ratios of 163, 334, and 426%, respectively, in the energy absorbed by the strengthened beams in comparison to the control beam. The shear strength of the strengthened beams was determined through analytical modeling by implementing a correction factor (α = 0.5) to take into consideration the debonding action between the SHCC layer and the beam sides. This factor significantly improved the predictive accuracy of the analytical models by matching the mean ratio of the analytical findings to the experimental predictions.
48
Jeong, Sugeun, Minseo Moon, and Daehyeon Kim. "Seismic Wave Amplification Characteristics in Slope Sections of Various Inclined Model Grounds." Applied Sciences 14, no. 19 (2024): 9014. http://dx.doi.org/10.3390/app14199014.
Full textAbstract:
The collapse of slopes caused by earthquakes can lead to landslides, resulting in significant damage to both lives and structures. Seismic reinforcement of these slopes can protect social systems during an earthquake. In South Korea, where more than 70% of the land is mountainous, the stability of slopes is of paramount importance compared to other countries. While many seismic designs are based on peak ground acceleration (PGA), there is relatively little consideration given to the extent of PGA’s influence, and few studies have been done. This study aims to assess the seismic amplification of slopes with multilayers using a 1 g shaking table and verify the results through numerical analysis after confirming the impact of PGA at specific points. Typically, slope model experiments are conducted on single-layered ground models. However, actual ground conditions consist of multiple layers rather than a single layer, so a multi-layered model was created with different properties for the upper and lower layers. Two multi-layered ground models consisting of two layers were created, one with a flat ground surface and the other with a sloped surface. The properties of the two layers in each model were configured as a single layer to create the slope models. The peak ground acceleration (PGA) of the four ground models was compared, revealing that seismic wave amplification increases as it moves upward, and the amplification is even greater when transitioning from the lower to the upper ground layers, leading to different dynamic behavior of the slope. Through the contour lines, the influence of PGA was further confirmed, and it was found that approximately 60% of the PGA impact occurs at the topmost part of the slope on average. Analysis of the earthquake waves showed that the top of the slope experienced an average amplification of about 31.75% compared to the input motion, while the lower part experienced an average amplification of about 27.85%. Numerical analysis was performed using the ABAQUS program, and the results were compared with the 1 g shaking table experiments through spectral acceleration (SA), showing good agreement with the experimental results.
49
Christiansen, Anders Vest, and Niels Bøie Christensen. "A quantitative appraisal of airborne and ground‐based transient electromagnetic (TEM) measurements in Denmark." GEOPHYSICS 68, no. 2 (2003): 523–34. http://dx.doi.org/10.1190/1.1567220.
Full textAbstract:
The last decade has seen growing use of ground‐based transient electromagnetic (TEM) methods in Denmark for hydrogeological purposes. Due to an intensified mapping campaign, airborne TEM methods were proposed as a possible tool for mapping large areas. The first test flights were flown in June 2000 using the GEOTEM system. Traditional approximate interpretation tools for airborne data are insufficient in hydrogeological investigations where a quantitative model specifying model parameter reliability is needed. We have carried out full nonlinear one‐dimensional inversion on the field amplitude of airborne synthetic and field data and compared the airborne method with the traditional ground‐based PROTEM 47 system that has found extensive use in Denmark. An improved measuring procedure for airborne systems is suggested to facilitate the estimation of noise that is necessary in a quantitative inversion. The analyses of synthetic data demonstrate the differences in resolution capability between ground‐based and airborne data. Ground‐based data typically resolve three‐ or four‐layer models and occasionally up to five layers. Airborne data resolve three layers as a maximum, one or two layers being common. The airborne GEOTEM system detects layers to depths of more than 300 m, bearing only little information about the top 50–70 m. The ground‐based PROTEM 47 system has a maximum penetration of approximately 170 m, with higher resolution capabilities in the top 100 m. Coupling to man‐made conductors is a serious problem for all TEM methods in densely populated areas and results in distorted data. Coupling influences the airborne data from Denmark on two‐thirds of the area covered. These data must be eliminated to avoid misinterpretation.
50
Zhang, Lanfei, Xu Chu, Siyu Ding, Mingshuo Zhou, Chenxu Ni, and Xingjian Wang. "Surrogate Modeling of Hydrogen-Enriched Combustion Using Autoencoder-Based Dimensionality Reduction." Processes 13, no. 4 (2025): 1093. https://doi.org/10.3390/pr13041093.
Full textAbstract:
Deep learning-based surrogate models have received wide attention for efficient and cost-effective predictions of fluid flows and combustion, while their hyperparameter settings often lack generalizable guidelines. This study examines two different types of surrogate models, convolutional autoencoder (CAE)-based reduced order models (ROMs) and fully connected autoencoder (FCAE)-based ROMs, for emulating hydrogen-enriched combustion from a triple-coaxial nozzle jet. The performances of these ROMs are discussed in detail, with an emphasis on key hyperparameters, including the number of network layers in the encoder (l), latent vector dimensionality (dim), and convolutional stride (s). The results indicate that a larger l is essential for capturing features in strongly nonlinear flowfields, whereas a smaller l is more effective for less nonlinear distributions, as additional layers may cause overfitting. Specifically, when employing CAE-based ROMs to predict the spatial distribution for H2 (XH2) with weak nonlinearity, the reconstruction absolute average relative deviation (AARD) from the two-layer model was marginally higher than that of three- and four-layer models, whereas the prediction AARD was approximately 5% lower. A smaller dim yields better performance in weakly nonlinear flowfields but may increase local errors in some cases due to excessive feature compression. A CAE-based ROM with a dim = 10 achieved a notably lower AARD of 4.01% for XH2 prediction. A smaller s may enhance the spatial resolution yet raise computational costs. Under identical hyperparameters, the CAE-based ROM outperformed the FCAE-based ROM in both cost-effectiveness and accuracy, achieving a 35 times faster training speed and lower absolute average relative deviation in prediction. These findings provide important guidelines for hyperparameter selection in training autoencoder (AE)-based ROMs for hydrogen-enriched combustion and other similar engineering design problems.