Dissertations / Theses on the topic 'Alignment of signals'

Thesis submitted in the fulfillment of the requirements for the Degree of Master in Biomedical Engineering
The increasing number of biomedical systems and applications for human body understanding creates a need for information extraction tools to use in biosignals. It’s important to comprehend the changes in the biosignal’s morphology over time, as they often contain critical information on the condition of the subject or the status of the experiment. The creation of tools that automatically analyze and extract relevant attributes from biosignals, providing important information to the user, has a significant value in the biosignal’s processing field. The present dissertation introduces new algorithms for time series clustering, where we are able to separate and organize unlabeled data into different groups whose signals are similar to each other. Signal processing algorithms were developed for the detection of a meanwave, which represents the signal’s morphology and behavior. The algorithm designed computes the meanwave by separating and averaging all cycles of a cyclic continuous signal. To increase the quality of information given by the meanwave, a set of wave-alignment techniques was also developed and its relevance was evaluated in a real database. To evaluate our algorithm’s applicability in time series clustering, a distance metric created with the information of the automatic meanwave was designed and its measurements were given as input to a K-Means clustering algorithm. With that purpose, we collected a series of data with two different modes in it. The produced algorithm successfully separates two modes in the collected data with 99.3% of efficiency. The results of this clustering procedure were compared to a mechanism widely used in this area, which models the data and uses the distance between its cepstral coefficients to measure the similarity between the time series.The algorithms were also validated in different study projects. These projects show the variety of contexts in which our algorithms have high applicability and are suitable answers to overcome the problems of exhaustive signal analysis and expert intervention. The algorithms produced are signal-independent, and therefore can be applied to any type of signal providing it is a cyclic signal. The fact that this approach doesn’t require any prior information and the preliminary good performance make these algorithms powerful tools for biosignals analysis and classification.

When applying statistical data analysis techniques to analytical chemical data, all variables must have correspondence over the samples dimension in order for the analysis to generate meaningful results. Peak shifts in NMR and chromatography destroys that correspondence and creates data matrices that have to be aligned before analysis. In this thesis, new methods are introduced that allow for automated transformation from unaligned raw data to aligned data matrices where each column corresponds to a unique signal. These methods are based around linear multivariate models for the peak shifts and Hough transform for establishing the parameters of these linear models. Methods for quantification under difficult conditions, such as crowded spectral regions, noisy data and unknown peak identities are also introduced. These methods include automated peak selection and a robust method for background subtraction. This thesis focuses on the processing of the data; the experimental work is secondary and is not discussed in great detail. All the developed methods are put together in a full procedure that takes us from raw data to a table of concentrations in a matter of minutes. The procedure is applied to 1H-NMR data from biological samples, which is one of the toughest alignment tasks available in the field of analytical chemistry. It is shown that the procedure performs consistently on the same level as much more labor intensive manual techniques such as Chenomx NMRSuite spectral profiling. Several kinds of datasets are evaluated using the procedure. Most of the data is from the field of Metabolomics, where the goal is to establish concentrations of as many small molecules as possible in biological samples.

High-Pressure, High-Temperature (HPHT) pressure gauges are commonly used in oil wells for pressure transient analysis. Mathematical models are used to relate input perturbation (e.g., flow rate transients) with output responses (e.g., pressure transients), and subsequently, solve an inverse problem that infers reservoir parameters. The indispensable use of pressure data in well testing motivates continued improvement in the accuracy (quality), sampling rate (quantity), and autonomy (lifetime) of pressure gauges. This body of work presents improvements in three areas of high-pressure, high-temperature quartz memory gauge technology: calibration accuracy, multi-tool signal alignment, and tool autonomy estimation. The discussion introduces the response surface methodology used to calibrate gauges, develops accuracy and autonomy estimates based on controlled tests, and where applicable, relies on field gauge drill stem test data to validate accuracy predictions. Specific contributions of this work include: - Application of the unpaired sample t-test, a first in quartz sensor calibration, which resulted in reduction of uncertainty in gauge metrology by a factor of 2.25, and an improvement in absolute and relative tool accuracies of 33% and 56%, accordingly. Greater accuracy yields more reliable data and a more sensitive characterization of well parameters. - Post-processing of measurements from 2+ tools using a dynamic time warp algorithm that mitigates gauge clock drifts. Where manual alignment methods account only for linear shifts, the dynamic algorithm elastically corrects nonlinear misalignments accumulated throughout a job with an accuracy that is limited only by the clock's time resolution. - Empirical modeling of tool autonomy based on gauge selection, battery pack, sampling mode, and average well temperature. A first of its kind, the model distills autonomy into two independent parameters, each a function of the same two orthogonal factors: battery power capacity and gauge current consumption as functions of sampling mode and well temperature -- a premise that, for 3+ gauge and battery models, reduces the design of future autonomy experiments by at least a factor of 1.5.

This thesis proposes specific signal processing and machine learning methodologies for automatically aligning the lyrics of a song to its corresponding audio recording. The research carried out falls in the broader field of music information retrieval (MIR) and in this respect, we aim at improving some existing state-of-the-art methodologies, by introducing domain-specific knowledge. The goal of this work is to devise models capable of tracking in the music audio signal the sequential aspect of one particular element of lyrics - the phonemes. Music can be understood as comprising different facets, one of which is lyrics. The models we build take into account the complementary context that exists around lyrics, which is any musical facet complementary to lyrics. The facets used in this thesis include the structure of the music composition, structure of a melodic phrase, the structure of a metrical cycle. From this perspective, we analyse not only the low-level acoustic characteristics, representing the timbre of the phonemes, but also higher-level characteristics, in which the complementary context manifests. We propose specific probabilistic models to represent how the transitions between consecutive sung phonemes are conditioned by different facets of complementary context. The complementary context, which we address, unfolds in time according to principles that are particular of a music tradition. To capture these, we created corpora and datasets for two music traditions, which have a rich set of such principles: Ottoman Turkish makam and Beijing opera. The datasets and the corpora comprise different data types: audio recordings, music scores, and metadata. From this perspective, the proposed models can take advantage both of the data and the music-domain knowledge of particular musical styles to improve existing baseline approaches. As a baseline, we choose a phonetic recognizer based on hidden Markov models (HMM): a widely-used methodology for tracking phonemes both in singing and speech processing problems. We present refinements in the typical steps of existing phonetic recognizer approaches, tailored towards the characteristics of the studied music traditions. On top of the refined baseline, we device probabilistic models, based on dynamic Bayesian networks (DBN) that represent the relation of phoneme transitions to its complementary context. Two separate models are built for two granularities of complementary context: the structure of a melodic phrase (higher-level) and the structure of the metrical cycle (finer-level). In one model we exploit the fact the syllable durations depend on their position within a melodic phrase. Information about the melodic phrases is obtained from the score, as well as from music-specific knowledge.Then in another model, we analyse how vocal note onsets, estimated from audio recordings, influence the transitions between consecutive vowels and consonants. We also propose how to detect the time positions of vocal note onsets in melodic phrases by tracking simultaneously the positions in a metrical cycle (i.e. metrical accents). In order to evaluate the potential of the proposed models, we use the lyrics-to-audio alignment as a concrete task. Each model improves the alignment accuracy, compared to the baseline, which is based solely on the acoustics of the phonetic timbre. This validates our hypothesis that knowledge of complementary context is an important stepping stone for computationally tracking lyrics, especially in the challenging case of singing with instrumental accompaniment. The outcomes of this study are not only theoretic methodologies and data, but also specific software tools that have been integrated into Dunya - a suite of tools, built in the context of CompMusic, a project for advancing the computational analysis of the world's music. With this application, we have also shown that the developed methodologies are useful not only for tracking lyrics, but also for other use cases, such as enriched music listening and appreciation, or for educational purposes.
La tesi aquí presentada proposa metodologies d’aprenentatge automàtic i processament de senyal per alinear automàticament el text d’una cançó amb el seu corresponent enregistrament d’àudio. La recerca duta a terme s’engloba en l’ampli camp de l’extracció d’informació musical (Music Information Retrieval o MIR). Dins aquest context la tesi pretén millorar algunes de les metodologies d’última generació del camp introduint coneixement específic de l’àmbit. L’objectiu d’aquest treball és dissenyar models que siguin capaços de detectar en la senyal d’àudio l’aspecte seqüencial d’un element particular dels textos musicals; els fonemes. Podem entendre la música com la composició de diversos elements entre els quals podem trobar el text. Els models que construïm tenen en compte el context complementari del text. El context són tots aquells aspectes musicals que complementen el text, dels quals hem utilitzat en aquest tesi: la estructura de la composició musical, la estructura de les frases melòdiques i els accents rítmics. Des d’aquesta prespectiva analitzem no només les característiques acústiques de baix nivell, que representen el timbre musical dels fonemes, sinó també les característiques d’alt nivell en les quals es fa patent el context complementari. En aquest treball proposem models probabilístics específics que representen com les transicions entre fonemes consecutius de veu cantanda es veuen afectats per diversos aspectes del context complementari. El context complementari que tractem aquí es desenvolupa en el temps en funció de les característiques particulars de cada tradició musical. Per tal de modelar aquestes característiques hem creat corpus i conjunts de dades de dues tradicions musicals que presenten una gran riquesa en aquest aspectes; la música de l’opera de Beijing i la música makam turc-otomana. Les dades són de diversos tipus; enregistraments d’àudio, partitures musicals i metadades. Des d’aquesta prespectiva els models proposats poden aprofitar-se tant de les dades en si mateixes com del coneixement específic de la tradició musical per a millorar els resultats de referència actuals. Com a resultat de referència prenem un reconeixedor de fonemes basat en models ocults de Markov (Hidden Markov Models o HMM), una metodologia abastament emprada per a detectar fonemes tant en la veu cantada com en la parlada. Presentem millores en els processos comuns dels reconeixedors de fonemes actuals, ajustant-los a les característiques de les tradicions musicals estudiades. A més de millorar els resultats de referència també dissenyem models probabilistics basats en xarxes dinàmiques de Bayes (Dynamic Bayesian Networks o DBN) que respresenten la relació entre la transició dels fonemes i el context complementari. Hem creat dos models diferents per dos aspectes del context complementari; la estructura de la frase melòdica (alt nivell) i la estructura mètrica (nivell subtil). En un dels models explotem el fet que la duració de les síl·labes depén de la seva posició en la frase melòdica. Obtenim aquesta informació sobre les frases musical de la partitura i del coneixement específic de la tradició musical. En l’altre model analitzem com els atacs de les notes vocals, estimats directament dels enregistraments d’àudio, influencien les transicions entre vocals i consonants consecutives. A més també proposem com detectar les posicions temporals dels atacs de les notes en les frases melòdiques a base de localitzar simultàniament els accents en un cicle mètric musical. Per tal d’evaluar el potencial dels mètodes proposats utlitzem la tasca específica d’alineament de text amb àudio. Cada model proposat millora la precisió de l’alineament en comparació als resultats de referència, que es basen exclusivament en les característiques acústiques tímbriques dels fonemes. D’aquesta manera validem la nostra hipòtesi de que el coneixement del context complementari ajuda a la detecció automàtica de text musical, especialment en el cas de veu cantada amb acompanyament instrumental. Els resultats d’aquest treball no consisteixen només en metodologies teòriques i dades, sinó també en eines programàtiques específiques que han sigut integrades a Dunya, un paquet d’eines creat en el context del projecte de recerca CompMusic, l’objectiu del qual és promoure l’anàlisi computacional de les músiques del món. Gràcies a aquestes eines demostrem també que les metodologies desenvolupades es poden fer servir per a altres aplicacions en el context de la educació musical o la escolta musical enriquida.

Cette thèse porte sur la création d'un système expert d'alignement automatique des signaux chromatographiques répondant à une problématique de dérives et de décalages de signaux rencontrée dans l'inter-comparaison de données en milieu évolutif. Après un état de l'art des différentes méthodes d'alignement qui existent dans la littérature, les performances des méthodes librement disponibles ont été testées sur des jeux de données chromatographiques simulées et réelles. A l'issu de ce travail méthodique, il s'est avéré qu'aucune des méthodes n'apportait pleinement satisfaction en matière de performances définies dans le cahier des charges. Ainsi, une optimisation de la meilleure de ces méthodes d'alignement a été développée afin qu'elle puisse être annexée à un logiciel d'acquisition et de traitement de données chromatographiques. La dernière partie de ce manuscrit traite d'une problématique complémentaire, la conformité des échantillons en terme de contrôle qualité. La similitude des pics est évaluée selon des critères développés et validés par une exploitation manuelle des données
This thesis focuses on the creation of an expert system for automatic alignment of chromatographic signals in response to a problem of drifts and shifts of signals encountered in the inter-comparison of data in evolving environment. After a state of the art of the different alignment methods that exist in the literature, the performances of freely available methods were tested on sets of simulated and real chromatographic data. At the end of this methodical work, it turned out that none of the methods did not provide fully satisfactory in terms of performance defined in the specification. Thus, an optimization of the best alignment method has been developed so that it can be attached to a software acquisition and processing of chromatographic data. The last part of this thesis deals with a complementary problem, the conformity of the samples in terms of quality control. The similarity of the peaks is evaluated according to criteria developed and validated by manual operation data

Introduction
Current instrumentation makes it possible to measure vehicle emissions with high temporal resolution. But the increased resolution of emissions signals does not equate with increased accuracy. A prerequisite for the derivation of accurate emission factors from instantaneous vehicle emissions data is a fine allocation of measured mass emissions to recorded engine or vehicle states. This poses a technical challenge, because vehicle emission test facilities are not designed to support instantaneous emissions modelling, and they introduce distorting effects that compromise the instantaneous accuracy of the measured signals.

Methodology
These distorting effects can be compensated through a combination of physical modelling and data post-processing. The main original contribution of this dissertation is a novel methodology for the compensation of instantaneous emission signals, which is fully described herein. Whereas previous methodologies relied on systems theory modelling, and on comprehensive testing to model the sub-systems of the measurement setup, the alternative approach uses CO2 as a tracer of the distortions brought about by the measurement setup, which is modelled as a 'lump' system.

Conclusions
The main benefits of this methodology are its low burden of experimental work and its flexibility. Furthermore, it has been fully implemented in the 'esto' software tool, which can perform the compensation of emission signals with minimal user intervention and speed up the creation of engine emission maps.

Dans cette thèse, nous étudions des modèles semi-paramétriques dits de forme invariante. Ces modèles consistent en l'observation d'un nombre fixés de fonctions de régression identiques à un opérateur de déformation paramétriques près. Ce type de modèles trouve des applications dans les problèmes d'alignement de signaux continus (images 2D, rythmes biologiques, ...) ou discrets (electroencéphalogramme, ...). Pour différents groupes de déformations, nous proposons des M-estimateurs pour les paramètres caractérisant les opérateurs associés aux fonctions de régression. Ces estimateurs minimisent ou maximisent des fonctions de contraste, construites à partir de la moyenne synchronisée des transformées de Fourier des données. De plus, pour l'un des modèles étudiés, nous prouvons l'efficacité semi-paramétrique de cet estimateur ainsi défini, et nous proposons un test d'adéquation du modèle de forme invariante construit à partir d'une des fonctions de contraste.

Cette thèse porte sur l'alignement automatique d'un enregistrement audio avec la partition de musique correspondante. Nous adoptons une approche probabiliste et proposons une démarche théorique pour la modélisation algorithmique de ce problème d'alignement automatique. La question est de modéliser l'évolution temporelle des événements par des processus stochastiques. Notre démarche part d'une spécificité de l'alignement musical : une partition attribue à chaque événement une durée nominale, qui est une information a priori sur la durée probable d'occurrence de l'événement. La problématique qui nous occupe est celle de la modélisation probabiliste de cette information de durée. Nous définissons la notion de cohérence temporelle à travers plusieurs critères de cohérence que devrait respecter tout algorithme d'alignement musical. Ensuite, nous menons une démarche axiomatique autour du cas des modèles de semi-Markov cachés. Nous démontrons que ces critères sont respectés lorsque des conditions mathématiques particulières sont vérifiées par les lois a priori du modèle probabiliste de la partition. Ces conditions proviennent de deux domaines mathématiques jusqu'ici étrangers à la question de l'alignement : les processus de Lévy et la totale positivité d'ordre deux. De nouveaux résultats théoriques sont démontrés sur l'interrelation entre ces deux notions. En outre, les bienfaits pratiques de ces résultats théoriques sont démontrés expérimentalement sur des algorithmes d'alignement en temps réel
This thesis deals with automatic alignment of audio recordings with corresponding music scores. We study algorithmic solutions for this problem in the framework of probabilistic models which represent hidden evolution on the music score as stochastic process. We begin this work by investigating theoretical foundations of the design of such models. To do so, we undertake an axiomatic approach which is based on an application peculiarity: music scores provide nominal duration for each event, which is a hint for the actual and unknown duration. Thus, modeling this specific temporal structure through stochastic processes is our main problematic. We define temporal coherency as compliance with such prior information and refine this abstract notion by stating two criteria of coherency. Focusing on hidden semi-Markov models, we demonstrate that coherency is guaranteed by specific mathematical conditions on the probabilistic design and that fulfilling these prescriptions significantly improves precision of alignment algorithms. Such conditions are derived by combining two fields of mathematics, Lévy processes and total positivity of order 2. This is why the second part of this work is a theoretical investigation which extends existing results in the related literature

The purpose of this project is to improve the existing version of the Sequential Alignment and Position Verification System (SAPVS) for functional proton radiosurgery and to evaluate its performance after improvement .

The Periodic Event Synchronization Unit aligns devices without the ability to be triggered by an external source. The primary function of the unit is to align the pattern trigger pulses of two pulse pattern generators which supply four inputs of a multiplexer. The pulse pattern generators lack the ability to start their code according to an external signal. When operating, the designed unit maintains a specific pattern alignment of two binary data streams of 5 gigabits per second as a multiplexer combines them into a data stream of four times the bit rate. In addition to alignment, the unit can introduce offsets of up to 50 nanoseconds to the pattern alignment which corresponds to 250 bits. The unit is designed to allow the alignment of other devices as well, requiring as input the two event signals of the same frequency which need to be aligned. In order to align the devices providing the event pulses, one of the devices must either accept an external clocking source or have the ability to frequency modulate the internal clock. In practice, the test system was able to achieve and maintain the desired signal characteristics from the output of the multiplexer. The unit's robust design is shown by providing alignment of patterns for the full operating range of the pulse pattern generators and allowing a generator pattern to be aligned to a generic event pulse. Use of multiple units allows alignment of additional devices. The development of the Periodic Event Synchronization Unit provided an inexpensive solution to creating very high bit rate signals using preexisting equipment, as no commercial products were found to accomplish the same function.

Předmětem této práce je vytvoření metody sloužící k identifikaci organismů z metagenomických dat. Doposud k tomuto účelu spolehlivě dostačovaly metody založené na zarovnání sekvencí s referenční databází. Množství dat ovšem s rozvojem sekvenačních technik rapidně roste a tyto metody se tak stávají díky své výpočetní náročnosti nevhodnými. V této diplomové práci je popsán postup nové techniky, která umožňuje klasifikaci metagenomických dat bez nutnosti zarovnání. Metoda spočívá v převedení sekvenovaných úseků na genomické signály ve formě fázových reprezentací, ze kterých jsou následně extrahovány vektory příznaků. Těmito příznaky jsou tři Hjorthovy deskriptory. Ty jsou dále vystaveny metodě maximalizace věrohodnosti směsi Gaussovských rozložení, která umožňuje spolehlivé roztřídění fragmentů podle jejich příslušnosti k organismu.

Background: Phosphorylation of proteins plays a crucial role in the regulation and activation of metabolic and signaling pathways and constitutes an important target for pharmaceutical intervention. Central to the phosphorylation process is the recognition of specific target sites by protein kinases followed by the covalent attachment of phosphate groups to the amino acids serine, threonine, or tyrosine. The experimental identification as well as computational prediction of phosphorylation sites (P-sites) has proved to be a challenging problem. Computational methods have focused primarily on extracting predictive features from the local, one-dimensional sequence information surrounding phosphorylation sites. Results: We characterized the spatial context of phosphorylation sites and assessed its usability for improved phosphorylation site predictions. We identified 750 non-redundant, experimentally verified sites with three-dimensional (3D) structural information available in the protein data bank (PDB) and grouped them according to their respective kinase family. We studied the spatial distribution of amino acids around phosphorserines, phosphothreonines, and phosphotyrosines to extract signature 3D-profiles. Characteristic spatial distributions of amino acid residue types around phosphorylation sites were indeed discernable, especially when kinase-family-specific target sites were analyzed. To test the added value of using spatial information for the computational prediction of phosphorylation sites, Support Vector Machines were applied using both sequence as well as structural information. When compared to sequence-only based prediction methods, a small but consistent performance improvement was obtained when the prediction was informed by 3D-context information. Conclusion: While local one-dimensional amino acid sequence information was observed to harbor most of the discriminatory power, spatial context information was identified as relevant for the recognition of kinases and their cognate target sites and can be used for an improved prediction of phosphorylation sites. A web-based service (Phos3D) implementing the developed structurebased P-site prediction method has been made available at http://phos3d.mpimp-golm.mpg.de.

The recent uprise of end-to-end speech translation models requires a new generation of parallel corpora, composed of a large amount of source language speech utterances aligned with their target language textual translations. We hereby show a pipeline and a set of methods to collect hundreds of hours of English audio-book recordings and align them with their Italian textual translations, using exclusively public domain resources gathered semi-automatically from the web. The pipeline consists in three main areas: text collection, bilingual text alignment, and forced alignment. For the text collection task, we show how to automatically find e-book titles in a target language by using machine translation, web information retrieval, and named entity recognition and translation techniques. For the bilingual text alignment task, we investigated three methods: the Gale–Church algorithm in conjunction with a small-size hand-crafted bilingual dictionary, the Gale–Church algorithm in conjunction with a bigger bilingual dictionary automatically inferred through statistical machine translation, and bilingual text alignment by computing the vector similarity of multilingual embeddings of concatenation of consecutive sentences. Our findings seem to indicate that the consecutive-sentence-embeddings similarity computation approach manages to improve the alignment of difficult sentences by indirectly performing sentence re-segmentation. For the forced alignment task, we give a theoretical overview of the preferred method depending on the properties of the text to be aligned with the audio, suggesting and using a TTS-DTW (text-to-speech and dynamic time warping) based approach in our pipeline. The result of our experiments is a publicly available multi-modal corpus composed of about 130 hours of English speech aligned with its Italian textual translation and split in 60561 triplets of English audio, English transcript, and Italian textual translation. We also post-processed the corpus so as to extract 40-MFCCs features from the audio segments and released them as a data-set.

Protein sequencing has produced overwhelming amount of protein sequences, especially in the last decade. Nevertheless, the majority of the proteins' functional and structural classes are still unknown, and experimental methods currently used to determine these properties are very expensive, laborious and time consuming. Therefore, automated computational methods are urgently required to accurately and reliably predict functional and structural classes of the proteins. Several bioinformatics methods have been developed to determine such properties of the proteins directly from their sequence information. Such methods that involve signal processing methods have recently become popular in the bioinformatics area and been investigated for the analysis of DNA and protein sequences and shown to be useful and generally help better characterise the sequences. However, there are various technical issues that need to be addressed in order to overcome problems associated with the signal processing methods for the analysis of the proteins sequences. Amino acid indices that are used to transform the protein sequences into signals have various applications and can represent diverse features of the protein sequences and amino acids. As the majority of indices have similar features, this project proposes a new set of computationally derived indices that better represent the original group of indices. A study is also carried out that resulted in finding a unique and universal set of best discriminating amino acid indices for the characterisation of allergenic proteins. This analysis extracts features directly from the protein sequences by using Discrete Fourier Transform (DFT) to build a classification model based on Support Vector Machines (SVM) for the allergenic proteins. The proposed predictive model yields a higher and more reliable accuracy than those of the existing methods. A new method is proposed for performing a multiple sequence alignment. For this method, DFT-based method is used to construct a new distance matrix in combination with multiple amino acid indices that were used to encode protein sequences into numerical sequences. Additionally, a new type of substitution matrix is proposed where the physicochemical similarities between any given amino acids is calculated. These similarities were calculated based on the 25 amino acids indices selected, where each one represents a unique biological protein feature. The proposed multiple sequence alignment method yields a better and more reliable alignment than the existing methods. In order to evaluate complex information that is generated as a result of DFT, Complex Informational Spectrum Analysis (CISA) is developed and presented. As the results show, when protein classes present similarities or differences according to the Common Frequency Peak (CFP) in specific amino acid indices, then it is probable that these classes are related to the protein feature that the specific amino acid represents. By using only the absolute spectrum in the analysis of protein sequences using the informational spectrum analysis is proven to be insufficient, as biologically related features can appear individually either in the real or the imaginary spectrum. This is successfully demonstrated over the analysis of influenza neuraminidase protein sequences. Upon identification of a new protein, it is important to single out amino acid responsible for the structural and functional classification of the protein, as well as the amino acids contributing to the protein's specific biological characterisation. In this work, a novel approach is presented to identify and quantify the relationship between individual amino acids and the protein. This is successfully demonstrated over the analysis of influenza neuraminidase protein sequences. Characterisation and identification problem of the Influenza A virus protein sequences is tackled through a Subgroup Discovery (SD) algorithm, which can provide ancillary knowledge to the experts. The main objective of the case study was to derive interpretable knowledge for the influenza A virus problem and to consequently better describe the relationships between subtypes of this virus. Finally, by using DFT-based sequence-driven features a Support Vector Machine (SVM)-based classification model was built and tested, that yields higher predictive accuracy than that of SD. The methods developed and presented in this study yield promising results and can be easily applied to proteomic fields.

L’alignement d’images paramétrique a de nombreuses applications pour la réalité augmentée, la compression vidéo ou encore le suivi d’objets. Dans cette thèse, nous nous intéressons notamment aux techniques de recalage d’images (template matching) reposant sur l’optimisation locale d’une fonctionnelle d’erreur. Ces approches ont conduit ces dernières années à de nombreux algorithmes efficaces pour le suivi d’objets. Cependant, les performances de ces algorithmes ont été peu étudiées lorsque les images sont dégradées par un bruit important comme c’est le cas, par exemple, pour des captures réalisées dans des conditions de faible luminosité. Dans cette thèse, nous proposons un nouveau formalisme, appelé formalisme bidirectionnel, qui unifie plusieurs approches de l’état de l’art. Ce formalisme est utilisé dans un premier temps pour porter un éclairage nouveau sur un grand nombre d’approches de la littérature et en particulier sur l’algorithme ESM (Efficient Second-order Minimization). Nous proposons ensuite une étude théorique approfondie de l’influence du bruit sur le processus d’alignement. Cette étude conduit à la définition de deux nouvelles familles d’algorithmes, les approches ACL (Asymmetric Composition on Lie Groups) et BCL (Bidirectional Composition on Lie Groups) qui permettent d’améliorer les performances en présence de niveaux de bruit asymétriques (Rapport Signal sur Bruit différent dans les images). L’ensemble des approches introduites sont validées sur des données synthétiques et sur des données réelles capturées dans des conditions de faible luminosité
Parametric image alignment is a fundamental task of many vision applications such as object tracking, image mosaicking, video compression and augmented reality. To recover the motion parameters, direct image alignment works by optimizing a pixel-based difference measure between a moving image and a fixed-image called template. In the last decade, many efficient algorithms have been proposed for parametric object tracking. However, those approaches have not been evaluated for aligning images of low SNR (Signal to Noise ratio) such as images captured in low-light conditions. In this thesis, we propose a new formulation of image alignment called Bidirectional Framework for unifying existing state of the art algorithms. First, this framework allows us to produce new insights on existing approaches and in particular on the ESM (Efficient Second-order Minimization) algorithm. Subsequently, we provide a theoretical analysis of image noise on the alignment process. This yields the definition of two new approaches : the ACL (Asymmetric Composition on Lie Groups) algorithm and the BCL (Bidirectional Composition on Lie Groups) algorithm, which outperform existing approaches in presence of images of different SNR. Finally, experiments on synthetic and real images captured under low-light conditions allow to evaluate the new and existing approaches under various noise conditions

Els instruments musicals digitals es descomponen usualment en dues parts: la interfície d'usuari i el motor de síntesi. Tradicionalment la interfície d'usuari pren el nom de controlador musical. L'objectiu d'aquesta tesi és el disseny d'un interfície que permeti el control de la síntesi de sons instrumentals a partir de la veu cantada.

Amb la present recerca, intentem relacionar la veu amb el so dels instruments musicals, tenint en compte tan la descripció del senyal de veu, com les corresponents estratègies de mapeig per un control adequat del sintetitzador.
Proposem dos enfocaments diferents, d'una banda el control d'un sintetitzador de veu cantada, i d'altra banda el control de la síntesi de sons instrumentals. Per aquest últim, suggerim una representació del senyal de veu com a gests vocals, que inclou una sèrie d'algoritmes d'anàlisis de veu. A la vegada, per demostrar els resultats obtinguts, hem desenvolupat dos prototips a temps real.
Los instrumentos musicales digitales se pueden separar en dos componentes: el interfaz de usuario y el motor de sintesis. El interfaz de usuario se ha denominado tradicionalmente controlador musical. El objectivo de esta tesis es el diseño de un interfaz que permita el control de la sintesis de sonidos instrumentales a partir de la voz cantada.

La presente investigación pretende relacionar las caracteristicas de la voz con el sonido de los instrumentos musicales, teniendo en cuenta la descripción de la señal de voz, como las correspondientes estrategias de mapeo para un control apropiado del sintetizador. Se proponen dos enfoques distintos, el control de un sintetizador de voz cantada, y el control de la sintesis de sonidos insturmentales. Para este último, se sugiere una representación de la señal de voz como gestos vocales, incluyendo varios algoritmos de analisis de voz. Los resultados obtenidos se demuestran con dos prototipos a tiempo real.
Digital musical instruments are usually decomposed in two main constituent parts: a user interface and a sound synthesis engine. The user interface is popularly referred as a musical controller, and its design is the primary objective of this dissertation. Under the title of singing-driven interfaces, we aim to design systems that allow controlling the synthesis of musical instruments sounds with the singing voice.

This dissertation searches for the relationships between the voice and the sound of musical instruments by addressing both, the voice signal description, as well as the mapping strategies for a meaningful control of the synthesized sound.
We propose two different approaches, one for controlling a singing voice synthesizer, and another for controlling the synthesis of instrumental sounds. For the latter, we suggest to represent voice signal as vocal gestures, contributing with several voice analysis methods.
To demonstrate the obtained results, we developed two real-time prototypes.

This thesis addresses several shortcomings on the current state of the art methodologies in music information retrieval (MIR). In particular, it proposes several computational approaches to automatically analyze and describe music scores and audio recordings of Ottoman-Turkish makam music (OTMM). The main contributions of the thesis are the music corpus that has been created to carry out the research and the audio-score alignment methodology developed for the analysis of the corpus. In addition, several novel computational analysis methodologies are presented in the context of common MIR tasks of relevance for OTMM. Some example tasks are predominant melody extraction, tonic identification, tempo estimation, makam recognition, tuning analysis, structural analysis and melodic progression analysis. These methodologies become a part of a complete system called Dunya-makam for the exploration of large corpora of OTMM. The thesis starts by presenting the created CompMusic Ottoman- Turkish makam music corpus. The corpus includes 2200 music scores, more than 6500 audio recordings, and accompanying metadata. The data has been collected, annotated and curated with the help of music experts. Using criteria such as completeness, coverage and quality, we validate the corpus and show its research potential. In fact, our corpus is the largest and most representative resource of OTMM that can be used for computational research. Several test datasets have also been created from the corpus to develop and evaluate the specific methodologies proposed for different computational tasks addressed in the thesis. The part focusing on the analysis of music scores is centered on phrase and section level structural analysis. Phrase boundaries are automatically identified using an existing state-of-the-art segmentation methodology. Section boundaries are extracted using heuristics specific to the formatting of the music scores. Subsequently, a novel method based on graph analysis is used to establish similarities across these structural elements in terms of melody and lyrics, and to label the relations semiotically. The audio analysis section of the thesis reviews the state-of-the-art for analysing the melodic aspects of performances of OTMM. It proposes adaptations of existing predominant melody extraction methods tailored to OTMM. It also presents improvements over pitch-distribution-based tonic identification and makam recognition methodologies. The audio-score alignment methodology is the core of the thesis. It addresses the culture-specific challenges posed by the musical characteristics, music theory related representations and oral praxis of OTMM. Based on several techniques such as subsequence dynamic time warping, Hough transform and variable-length Markov models, the audio-score alignment methodology is designed to handle the structural differences between music scores and audio recordings. The method is robust to the presence of non-notated melodic expressions, tempo deviations within the music performances, and differences in tonic and tuning. The methodology utilizes the outputs of the score and audio analysis, and links the audio and the symbolic data. In addition, the alignment methodology is used to obtain score-informed description of audio recordings. The scoreinformed audio analysis not only simplifies the audio feature extraction steps that would require sophisticated audio processing approaches, but also substantially improves the performance compared with results obtained from the state-of-the-art methods solely relying on audio data. The analysis methodologies presented in the thesis are applied to the CompMusic Ottoman-Turkish makam music corpus and integrated into a web application aimed at culture-aware music discovery. Some of the methodologies have already been applied to other music traditions such as Hindustani, Carnatic and Greek music. Following open research best practices, all the created data, software tools and analysis results are openly available. The methodologies, the tools and the corpus itself provide vast opportunities for future research in many fields such as music information retrieval, computational musicology and music education.
Esta tesis aborda varias limitaciones de las metodologías más avanzadas en el campo de recuperación de información musical (MIR por sus siglas en inglés). En particular, propone varios métodos computacionales para el análisis y la descripción automáticas de partituras y grabaciones de audio de música de makam turco-otomana (MMTO). Las principales contribuciones de la tesis son el corpus de música que ha sido creado para el desarrollo de la investigación y la metodología para alineamiento de audio y partitura desarrollada para el análisis del corpus. Además, se presentan varias metodologías nuevas para análisis computacional en el contexto de las tareas comunes de MIR que son relevantes para MMTO. Algunas de estas tareas son, por ejemplo, extracción de la melodía predominante, identificación de la tónica, estimación de tempo, reconocimiento de makam, análisis de afinación, análisis estructural y análisis de progresión melódica. Estas metodologías constituyen las partes de un sistema completo para la exploración de grandes corpus de MMTO llamado Dunya-makam. La tesis comienza presentando el corpus de música de makam turcootomana de CompMusic. El corpus incluye 2200 partituras, más de 6500 grabaciones de audio, y los metadatos correspondientes. Los datos han sido recopilados, anotados y revisados con la ayuda de expertos. Utilizando criterios como compleción, cobertura y calidad, validamos el corpus y mostramos su potencial para investigación. De hecho, nuestro corpus constituye el recurso de mayor tamaño y representatividad disponible para la investigación computacional de MMTO. Varios conjuntos de datos para experimentación han sido igualmente creados a partir del corpus, con el fin de desarrollar y evaluar las metodologías específicas propuestas para las diferentes tareas computacionales abordadas en la tesis. La parte dedicada al análisis de las partituras se centra en el análisis estructural a nivel de sección y de frase. Los márgenes de frase son identificados automáticamente usando uno de los métodos de segmentación existentes más avanzados. Los márgenes de sección son extraídos usando una heurística específica al formato de las partituras. A continuación, se emplea un método de nueva creación basado en análisis gráfico para establecer similitudes a través de estos elementos estructurales en cuanto a melodía y letra, así como para etiquetar relaciones semióticamente. La sección de análisis de audio de la tesis repasa el estado de la cuestión en cuanto a análisis de los aspectos melódicos en grabaciones de MMTO. Se proponen modificaciones de métodos existentes para extracción de melodía predominante para ajustarlas a MMTO. También se presentan mejoras de metodologías tanto para identificación de tónica basadas en distribución de alturas, como para reconocimiento de makam. La metodología para alineación de audio y partitura constituye el grueso de la tesis. Aborda los retos específicos de esta cultura según vienen determinados por las características musicales, las representaciones relacionadas con la teoría musical y la praxis oral de MMTO. Basada en varias técnicas tales como deformaciones dinámicas de tiempo subsecuentes, transformada de Hough y modelos de Markov de longitud variable, la metodología de alineamiento de audio y partitura está diseñada para tratar las diferencias estructurales entre partituras y grabaciones de audio. El método es robusto a la presencia de expresiones melódicas no anotadas, desviaciones de tiempo en las grabaciones, y diferencias de tónica y afinación. La metodología utiliza los resultados del análisis de partitura y audio para enlazar el audio y los datos simbólicos. Además, la metodología de alineación se usa para obtener una descripción informada por partitura de las grabaciones de audio. El análisis de audio informado por partitura no sólo simplifica los pasos para la extracción de características de audio que de otro modo requerirían sofisticados métodos de procesado de audio, sino que también mejora sustancialmente su rendimiento en comparación con los resultados obtenidos por los métodos más avanzados basados únicamente en datos de audio. Las metodologías analíticas presentadas en la tesis son aplicadas al corpus de música de makam turco-otomana de CompMusic e integradas en una aplicación web dedicada al descubrimiento culturalmente específico de música. Algunas de las metodologías ya han sido aplicadas a otras tradiciones musicales, como música indostaní, carnática y griega. Siguiendo las mejores prácticas de investigación en abierto, todos los datos creados, las herramientas de software y los resultados de análisis está disponibles públicamente. Las metodologías, las herramientas y el corpus en sí mismo ofrecen grandes oportunidades para investigaciones futuras en muchos campos tales como recuperación de información musical, musicología computacional y educación musical.
Aquesta tesi adreça diverses deficiències en l’estat actual de les metodologies d’extracció d’informació de música (Music Information Retrieval o MIR). En particular, la tesi proposa diverses estratègies per analitzar i descriure automàticament partitures musicals i enregistraments d’actuacions musicals de música Makam Turca Otomana (OTMM en les seves sigles en anglès). Les contribucions principals de la tesi són els corpus musicals que s’han creat en el context de la tesi per tal de dur a terme la recerca i la metodologia de alineament d’àudio amb la partitura que s’ha desenvolupat per tal d’analitzar els corpus. A més la tesi presenta diverses noves metodologies d’anàlisi computacional d’OTMM per a les tasques més habituals en MIR. Alguns exemples d’aquestes tasques són la extracció de la melodia principal, la identificació del to musical, l’estimació de tempo, el reconeixement de Makam, l’anàlisi de la afinació, l’anàlisi de la estructura musical i l’anàlisi de la progressió melòdica. Aquest seguit de metodologies formen part del sistema Dunya-makam per a la exploració de grans corpus musicals d’OTMM. En primer lloc, la tesi presenta el corpus CompMusic Ottoman- Turkish makam music. Aquest inclou 2200 partitures musicals, més de 6500 enregistraments d’àudio i metadata complementària. Les dades han sigut recopilades i anotades amb ajuda d’experts en aquest repertori musical. El corpus ha estat validat en termes de d’exhaustivitat, cobertura i qualitat i mostrem aquí el seu potencial per a la recerca. De fet, aquest corpus és el la font més gran i representativa de OTMM que pot ser utilitzada per recerca computacional. També s’han desenvolupat diversos subconjunts de dades per al desenvolupament i evaluació de les metodologies específiques proposades per a les diverses tasques computacionals que es presenten en aquest tesi. La secció de la tesi que tracta de l’anàlisi de partitures musicals se centra en l’anàlisi estructural a nivell de secció i de frase musical. Els límits temporals de les frases musicals s’identifiquen automàticament gràcies a un metodologia de segmentació d’última generació. Els límits de les seccions s’extreuen utilitzant un seguit de regles heurístiques determinades pel format de les partitures musicals. Posteriorment s’utilitza un nou mètode basat en anàlisi gràfic per establir semblances entre aquest elements estructurals en termes de melodia i text. També s’utilitza aquest mètode per etiquetar les relacions semiòtiques existents. La següent secció de la tesi tracta sobre anàlisi d’àudio i en particular revisa les tecnologies d’avantguardia d’anàlisi dels aspectes melòdics en OTMM. S’hi proposen adaptacions dels mètodes d’extracció de melodia existents que s’ajusten a OTMM. També s’hi presenten millores en metodologies de reconeixement de makam i en identificació de tònica basats en distribució de to. La metodologia d’alineament d’àudio amb partitura és el nucli de la tesi. Aquesta aborda els reptes culturalment específics imposats per les característiques musicals, les representacions de la teoria musical i la pràctica oral particulars de l’OTMM. Utilitzant diverses tècniques tal i com Dynamic Time Warping, Hough Transform o models de Markov de durada variable, la metodologia d’alineament esta dissenyada per enfrontar les diferències estructurals entre partitures musicals i enregistraments d’àudio. El mètode és robust inclús en presència d’expressions musicals no anotades en la partitura, desviacions de tempo ocorregudes en les actuacions musicals i diferències de tònica i afinació. La metodologia aprofita els resultats de l’anàlisi de la partitura i l’àudio per enllaçar la informació simbòlica amb l’àudio. A més, la tècnica d’alineament s’utilitza per obtenir descripcions de l’àudio fonamentades en la partitura. L’anàlisi de l’àudio fonamentat en la partitura no només simplifica les fases d’extracció de característiques d’àudio que requeririen de mètodes de processament d’àudio sofisticats, sinó que a més millora substancialment els resultats comparat amb altres mètodes d´ultima generació que només depenen de contingut d’àudio. Les metodologies d’anàlisi presentades s’han utilitzat per analitzar el corpus CompMusic Ottoman-Turkish makam music i s’han integrat en una aplicació web destinada al descobriment musical de tradicions culturals específiques. Algunes de les metodologies ja han sigut també aplicades a altres tradicions musicals com la Hindustani, la Carnàtica i la Grega. Seguint els preceptes de la investigació oberta totes les dades creades, eines computacionals i resultats dels anàlisis estan disponibles obertament. Tant les metodologies, les eines i el corpus en si mateix proporcionen àmplies oportunitats per recerques futures en diversos camps de recerca tal i com la musicologia computacional, la extracció d’informació musical i la educació musical. Traducció d’anglès a català per Oriol Romaní Picas.

La preorganització electrostàtica del centre actiu s'ha postulat com el mecanisme genèric de l'acció dels enzims. Així, alguns residus "estratègics" es disposarien per catalitzar reaccions interaccionant en una forma més forta amb l'estat de transició, baixant d'aquesta manera el valor de l'energia dactivació g cat. S'ha proposat que aquesta preorientació electrostática s'hauria de poder mostrar analitzant l'estabilitat electrostàtica de residus individuals en el centre actiu.
Ras es una proteïna essencial de senyalització i actúa com un interruptor cel.lular. Les característiques estructurals de Ras en el seu estat actiu (ON) són diferents de les que té a l'estat inactiu (OFF). En aquesta tesi es duu a terme una anàlisi exhaustiva de l'estabilitat dels residus del centre actiu deRas en l'estat actiu i inactiu.
The electrostatic preorganization of the active site has been put forward as the general framework of action of enzymes. Thus, enzymes would position "strategic" residues in such a way to be prepared to catalyze reactions by
interacting in a stronger way with the transition state, in this way decreasing the activation energy g cat for the catalytic process. It has been proposed that
such electrostatic preorientation should be shown by analyzing the electrostatic stability of individual residues in the active site.
Ras protein is an essential signaling molecule and functions as a switch in the
cell. The structural features of the Ras protein in its active state (ON state) are different than those in its inactive state (OFF state). In this thesis, an exhaustive analysis of the stability of residues in the active and inactive Ras active site is performed.

With the rapid advances in process technologies, the performance of state-of-the-art integrated circuits is improving steadily. The drive for higher performance is accompanied with increased emphasis on meeting timing constraints not only at the design phase but during device operation as well. Fortunately, technology advancements allow for even more precise control of the timing of digital signals, an advantage which can be used to provide solutions that can address some of the emerging timing issues. In this thesis, circuit and architectural techniques for the precise timing of digital signals are explored. These techniques are demonstrated in applications addressing timing issues in modern digital systems. A methodology for slow-speed timing characterization of high-speed pipelined datapaths is proposed. The technique uses a clock-timing circuit to create shifted versions of a slow-speed clock. These clocks control the data flow in the pipeline in the test mode. Test results show that the design provides an average timing resolution of 52.9ps in 0.18μm CMOS technology. Results also demonstrate the ability of the technique to track the performance of high-speed pipelines at a reduced clock frequency and to test the clock-timing circuit itself. In order to achieve higher resolutions than that of an inverter/buffer stage, a differential (vernier) delay line is commonly used. To allow for the design of differential delay lines with programmable delays, a digitally-controlled delay-element is proposed. The delay element is monotonic and achieves a high degree of transfer characteristics' (digital code vs. delay) linearity. Using the proposed delay element, a sub-1ps resolution is demonstrated experimentally in 0.18μm CMOS. The proposed delay element with a fixed delay step of 2ps is used to design a high-precision all-digital phase aligner. High-precision phase alignment has many applications in modern digital systems such as high-speed memory controllers, clock-deskew buffers, and delay and phase-locked loops. The design is based on a differential delay line and a variation tolerant phase detector using redundancy. Experimental results show that the phase aligner's range is from -264ps to +247ps which corresponds to an average delay step of approximately 2.43ps. For various input phase difference values, test results show that the difference is reduced to less than 2ps at the output of the phase aligner. On-chip time measurement is another application that requires precise timing. It has applications in modern automatic test equipment and on-chip characterization of jitter and skew. In order to achieve small conversion time, a flash time-to-digital converter is proposed. Mismatch between the various delay comparators limits the time measurement precision. This is demonstrated through an experiment in which a 6-bit, 2.5ps resolution flash time-to-digital converter provides an effective resolution of only 4-bits. The converter achieves a maximum conversion rate of 1.25GSa/s.

abstract: Genomic and proteomic sequences, which are in the form of deoxyribonucleic acid (DNA) and amino acids respectively, play a vital role in the structure, function and diversity of every living cell. As a result, various genomic and proteomic sequence processing methods have been proposed from diverse disciplines, including biology, chemistry, physics, computer science and electrical engineering. In particular, signal processing techniques were applied to the problems of sequence querying and alignment, that compare and classify regions of similarity in the sequences based on their composition. However, although current approaches obtain results that can be attributed to key biological properties, they require pre-processing and lack robustness to sequence repetitions. In addition, these approaches do not provide much support for efficiently querying sub-sequences, a process that is essential for tracking localized database matches. In this work, a query-based alignment method for biological sequences that maps sequences to time-domain waveforms before processing the waveforms for alignment in the time-frequency plane is first proposed. The mapping uses waveforms, such as time-domain Gaussian functions, with unique sequence representations in the time-frequency plane. The proposed alignment method employs a robust querying algorithm that utilizes a time-frequency signal expansion whose basis function is matched to the basic waveform in the mapped sequences. The resulting WAVEQuery approach is demonstrated for both DNA and protein sequences using the matching pursuit decomposition as the signal basis expansion. The alignment localization of WAVEQuery is specifically evaluated over repetitive database segments, and operable in real-time without pre-processing. It is demonstrated that WAVEQuery significantly outperforms the biological sequence alignment method BLAST for queries with repetitive segments for DNA sequences. A generalized version of the WAVEQuery approach with the metaplectic transform is also described for protein sequence structure prediction. For protein alignment, it is often necessary to not only compare the one-dimensional (1-D) primary sequence structure but also the secondary and tertiary three-dimensional (3-D) space structures. This is done after considering the conformations in the 3-D space due to the degrees of freedom of these structures. As a result, a novel directionality based 3-D waveform mapping for the 3-D protein structures is also proposed and it is used to compare protein structures using a matched filter approach. By incorporating a 3-D time axis, a highly-localized Gaussian-windowed chirp waveform is defined, and the amino acid information is mapped to the chirp parameters that are then directly used to obtain directionality in the 3-D space. This mapping is unique in that additional characteristic protein information such as hydrophobicity, that relates the sequence with the structure, can be added as another representation parameter. The additional parameter helps tracking similarities over local segments of the structure, this enabling classification of distantly related proteins which have partial structural similarities. This approach is successfully tested for pairwise alignments over full length structures, alignments over multiple structures to form a phylogenetic trees, and also alignments over local segments. Also, basic classification over protein structural classes using directional descriptors for the protein structure is performed.
Dissertation/Thesis
Ph.D. Electrical Engineering 2011

碩士
國立中央大學
通訊工程學系
102
Attracted by the benefits of multi-antenna relaying for next generation wireless communication in enhancing the system capacity, cellular coverage and spectral efficiency. While the previous schemes are propose on the transmitter and receiver cooperate design between base station and user nodes which achieves better performance, our research focus on developing the interference rejection and management technique for next generation cellular relay networks. For many research only focuses on the scheme of uplink communication, the downlink broadcasting transmission is rare to seen. In our study, we proposed a jointly design rules of uplink and downlink transmission technique for multi-user tow-way relay network through the block diagonalization broadcasting method joint with our propose multi-user receive antenna selection method. In this research, we consider an application jointly design the transmit precoder and signal alignment space selection for multi-user two-way relay network, provide an efficient data transmission method. The main contribution of this research is to provide a series of design rules including the user transmit precoder design, relay broadcasting beamforming weight design and the signal alignment space selection based on interference management and achievable sum-rate maximization for multi-user relay networks.

博士
國立成功大學
工程科學系碩博士班
94
Many data mining methods have recently been applied in a wide variety of fields. However, data collected for mining has many class-imbalance problems. These problems are difficult to solve using conventional probability and statistical methods, which usually find well-known rules and do not easily discover the patterns of interest in specific fields. For example, the critical risk patterns for credit scoring in finance and the function pattern for novel protein in molecular biology are both are difficult to find with general algorithms owing to class imbalance. In this study, one hybrid system was designed to effectively identify patterns of interest on imbalanced datasets, and developed another highly effective hybrid system to accurately predict protein functions in accordance with sequence alignment by signal registration. 　First, to effectively detect patterns of interest on an imbalanced dataset, this work applies extensively adopted data mining approaches to perform credit scoring tasks following the standard data mining procedure. Many experiments are also conducted with the banking data of real-world case, which comprise interesting patterns and imbalanced classes. These approaches include ten types of neural networks, C4.5, ID3, PRISM, NNge, IBk, Naive Bayes, Complement Naive Bayes, BayesNetB, Random Committee, and Voted Perceptron. Experimental results indicate that the above-mentioned methods have difficulty in prediction for the risk patterns of interest. The most significant factor in the difficulty regarding the prediction accuracy for interesting pattern is the ratio of amount of instances, which directly influence the judgments on contradiction. After performing the data cleaning step by combining the two above-mentioned methods, this study concludes that the combination of the PRISM method, which successfully classifies of non-contradiction instances as a filter, and the ID3 decision tree, which is easy to understand as a classifier, has the best total accuracy of prediction. However, the interesting patterns of minority class still cannot be effectively confirmed. Thus, this study develops the first hybrid system, an entropy-based neuro-fuzzy network with multiple decision trees quantifying the conflict-sensitivity contexture that denotes the index of contradictory decision-making. Experimental results reveal that the proposed method can effectively improve the prediction accuracy regarding interesting patterns in minority classes, while the accuracy of other classes also falls into the reasonable scopes. Additionally, this work describes many valuable business rules. 　Consequently, depending on the above-mentioned conclusions this following study addresses the theory in which the interesting patterns may be found in the imbalanced dataset of the real-world case, and may be aware of the change in space-time. Therefore, the second portion of this study designs another flexible hybrid system for signal registration, and applies it to predict the complex protein functions. The framework of the hybrid system is constructed by genetic programming, using Bayes causal tree as the data structure for individual representations. The system takes several protein sequences of known partial functions, and one targeted protein sequence of unknown functions, as the input. Next, the best causal tree for a local alignment of protein sequence to the multiple function classification is produced by simultaneous evolution with three fitness functions. The first fitness function is designed to evaluate similar features of the moment invariant into a set of signals, into which the fragments of a protein sequence are translated. The similar features are matched with one another by the robust point match (RPM) derived from the thin-plate spline theory with smoothing interpolation. The RPM performs the matches for geometric transformation to align signals classified in the same protein function with the variation of biochemistry environment, signal registration. The QR-decomposition in RPM is used to resolve the optimization of a mapping correspondence matrix. Furthermore, the RPM utilizes the one-to-many query and rooted one-way winner-take-all (RO-WTA) strategy to heap function nodes as the better Bayes causal tree based on the minimum difference between the known and unknown functions of protein. The second fitness function estimates the alignment score returned from the Smith-Waterman algorithm, which is an optimal sequence alignment method based on dynamic programming. These local sequence alignment scores for all nodes in the causal tree are added into a fitness value, and then the causal tree with the minimum value is obtained. The third fitness function estimates the coverage ranges in the light of the beginning and the end positions of known functions in real-world, and calculates the Bayes probability of causal tree so that the first and the second fitness functions are restricted to a reasonable range. Finally, the best causal tree is determined using an inter-population exchange strategy by immigrating and emigrating between the three populations with difference fitness functions. Each terminal node in the resulting causal tree is composed of a fragmented protein sequence of unknown function, while each internal node in the same tree involves with a fragmented protein sequence of known partial function. A set of functions in the path from the root node to each terminal node indicates an evolutionary motif for each fragmented protein sequence with an unknown function. 　The experimental results confirm that the classification of protein function to biodiversity caused by the change in space-time can be subdivided using signal registration. By appropriately controlling annealing temperature, the results demonstrate that the local sequence alignment, the feature matched by RPM, and the sampling of classification in the real world converge to a stable point. Thus, the proposed system may restore the unknown function to the original real-world classification of known protein function. 　Because the hybrid system requires considerable computation time when many long sequences are involved, it simultaneously distributes the evolutionary computation to many parallel processes. Although the step using the Smith-Waterman algorithm for sequence alignment is the most time-consuming part, this system not only separates a long sequence into many small fragmented sequences with a set of sequence positions according to locally minimum signals using wavelet reconstruction and differential equation, but also adopts the previous computation results as the same sequence alignment involved. Thus, the system can deal with thousand of amino acids, such as SARS virus sequences. 　In the future, if the predicted protein function can be confirmed after combining other ongoing studies, including the novel gene findings and the microarray-based gene expression experiments, then the methods and theorems proposed in this dissertation can be widely applied in Bioinformatics to help understand biochemical pathways in proteomics.

Sequences of feature vectors are a natural way of representing temporal data. Given a database of sequences, a fundamental task is to find the database entry which is the most similar to a query. In this thesis, we present learning-based methods for efficiently and accurately comparing sequences in order to facilitate large-scale sequence search. Throughout, we will focus on the problem of matching MIDI files (a digital score format) to a large collection of audio recordings of music. The combination of our proposed approaches enables us to create the largest corpus of paired MIDI files and audio recordings ever assembled. Dynamic time warping (DTW) has proven to be an extremely effective method for both aligning and matching sequences. However, its performance is heavily affected by factors such as the feature representation used and its adjustable parameters. We therefore investigate automatically optimizing DTW-based alignment and matching of MIDI and audio data. Our approach uses Bayesian optimization to tune system design and parameters over a synthetically-created dataset of audio and MIDI pairs. We then perform an exhaustive search over DTW score normalization techniques to find the optimal method for reporting a reliable alignment confidence score, as required in matching tasks. This results in a DTW-based system which is conceptually simple and highly accurate at both alignment and matching. We also verify that this system achieves high performance in a large-scale qualitative evaluation of real-world alignments. Unfortunately, DTW can be far too inefficient for large-scale search when sequences are very long and consist of high-dimensional feature vectors. We therefore propose a method for mapping sequences of continuously-valued feature vectors to downsampled sequences of binary vectors. Our approach involves training a pair of convolutional networks to map paired groups of subsequent feature vectors to a Hamming space where similarity is preserved. Evaluated on the task of matching MIDI files to a large database of audio recordings, we show that this technique enables 99.99\% of the database to be discarded with a modest false reject rate while only requiring 0.2\% of the time to compute. Even when sped-up with a more efficient representation, the quadratic complexity of DTW greatly hinders its feasibility for very large-scale search. This cost can be avoided by mapping entire sequences to fixed-length vectors in an embedded space where sequence similarity is approximated by Euclidean distance. To achieve this embedding, we propose a feed-forward attention-based neural network model which can integrate arbitrarily long sequences. We show that this approach can extremely efficiently prune 90\% of our audio recording database with high confidence. After developing these approaches, we applied them together to the practical task of matching 178,561 unique MIDI files to the Million Song Dataset. The resulting ``Lakh MIDI Dataset'' provides a potential bounty of ground truth information for audio content-based music information retrieval. This can include transcription, meter, lyrics, and high-level musicological features. The reliability of the resulting annotations depends both on the quality of the transcription and the accuracy of the score-to-audio alignment. We therefore establish a baseline of reliability for score-derived information for different content-based MIR tasks. Finally, we discuss potential future uses of our dataset and the learning-based sequence comparison methods we developed.