Tesis sobre el tema "參數估計"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores tesis para su investigación sobre el tema "參數估計".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore tesis sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
WENG, GUANG-HUI, та 翁光輝. "尺度參數比例之估計". Thesis, 1990. http://ndltd.ncl.edu.tw/handle/53682071672297524349.
Texto completo蔡文華. "混合模型下的參數估計". Thesis, 2007. http://ndltd.ncl.edu.tw/handle/72899285367842876120.
Texto completo徐碧珻. "Birnbaum-Saunders 分配中參數之估計". Thesis, 1986. http://ndltd.ncl.edu.tw/handle/49812887364557974303.
Texto completoZhou, Fang-Min, та 周芳敏. "CARCH模型族之半參數估計". Thesis, 1996. http://ndltd.ncl.edu.tw/handle/86401728320379493058.
Texto completo黃雯敏. "指數分配參數之估計量的比較". Thesis, 1992. http://ndltd.ncl.edu.tw/handle/07454612334835287663.
Texto completo林書綺. "幾何觀念的統計模型參數估計". Thesis, 2000. http://ndltd.ncl.edu.tw/handle/48654598380350016707.
Texto completoZHAN, CHONG-YOU, та 詹沖祐. "廣義混舍模式參數之估計". Thesis, 1992. http://ndltd.ncl.edu.tw/handle/28573979241469199533.
Texto completo陳佳典. "FEXP模型之參數估計與推論". Thesis, 2005. http://ndltd.ncl.edu.tw/handle/35538731182926595700.
Texto completoKang, Wei, та 康瑋. "多項式分配之參數區間估計". Thesis, 2009. http://ndltd.ncl.edu.tw/handle/78918719178291449411.
Texto completo林子淵. "位置參數的穩健貝氏估計式". Thesis, 2001. http://ndltd.ncl.edu.tw/handle/79791682174449148018.
Texto completoJan, Hau-Wen, та 詹皓文. "LOTUS方法下參數區間估計方法". Thesis, 2015. http://ndltd.ncl.edu.tw/handle/72x968.
Texto completoXU, GUO-JUN, та 徐國鈞. "固定床反應系統之參數估計". Thesis, 1992. http://ndltd.ncl.edu.tw/handle/91833512355632164782.
Texto completoChang, Chaio-Hsiang, та 張喬翔. "Weibull比例風險模型的參數估計". Thesis, 2000. http://ndltd.ncl.edu.tw/handle/38512553621039750405.
Texto completoZHENG, SHUN-LIN, та 鄭順林. "多維連續指數族的參數改進估計式". Thesis, 1989. http://ndltd.ncl.edu.tw/handle/49957603451745684531.
Texto completoYang, Chengen, та 楊承恩. "GEV模型及其參數估計法之研究". Thesis, 2011. http://ndltd.ncl.edu.tw/handle/53853963974765831022.
Texto completo黃雅琪. "時間序列迴歸參數之穩健估計". Thesis, 2005. http://ndltd.ncl.edu.tw/handle/93209267146464120421.
Texto completoCHEN, JIN-XIANG, та 陳景祥. "用成群資料作韋氏參數的估計". Thesis, 1986. http://ndltd.ncl.edu.tw/handle/41289237676861117697.
Texto completoLai, Sheng-Hsiang, та 賴聲翔. "MOB 迴歸樹的參數區間估計方法". Thesis, 2015. http://ndltd.ncl.edu.tw/handle/p69223.
Texto completo徐銘嶽. "韋伯雙邊設限資料的參數估計". Thesis, 2000. http://ndltd.ncl.edu.tw/handle/24261892785036775688.
Texto completoLI, XIU-MEI, та 李秀美. "加速壽命測試下的參數估計法". Thesis, 1988. http://ndltd.ncl.edu.tw/handle/59300493080923894790.
Texto completoChen, Guan-Yu, та 陳冠宇. "在不同環境下估計計量遺傳參數之設計規劃". Thesis, 1995. http://ndltd.ncl.edu.tw/handle/99098931488895262308.
Texto completoChing, Wang Wen, та 王文卿. "DINA模式與G-DINA模式參數估計比較". Thesis, 2010. http://ndltd.ncl.edu.tw/handle/13735434579616502705.
Texto completo國立臺中教育大學
教育測驗統計研究所
98
Demands of today's education reform is ─ "No Child Left Behind." But the prerequisite is to know the strengths and weaknesses of each student. Then we can design the remedial program. The traditional test only shows the students’ scale score in the groups, but can’t show the message of whether a student mastery a skill, and can’t help the students or the teachers a better understanding of the meaning represented by scores, and make more efficient learning. The cognitive diagnosis models can provide a solution. In the cognitive diagnostic models, DINA model is the simplest and easiest to explain. Many scholars have been involved in the application of research, model development, etc. This study simulate the Q matrix with different design on variety parameters in items, amounts of attributes , sample sizes, respectively. And uses DINA and G-DINA model to estimate the effectiveness of the differences. The results showed: 1. When Q matrix is unbalance, the accuracy of estimation is better. 2. When sample’s size increased, it may improve the accuracy of estimation. 3. When the item’s parameter value is smaller, the accuracy of estimation is higher. 4. When amount of items is the same, the amount of attributes is larger, the accuracy of estimation is better. 5. When the amount of attributes is the same, the test’s length is larger, the accuracy of estimation is better. 6. Simulated data estimated by DINA model is better than G-DINA model. 7. The empirical data estimated by DINA model and G-DINA model has the same results with this study.
CHEN, GUO-JIE, та 陳國杰. "加權參數估計在多變數系統的控制及應用". Thesis, 1986. http://ndltd.ncl.edu.tw/handle/97581003071579125223.
Texto completoKuo, Chun-Yi, та 郭春益. "型Ⅱ設限資料可靠度參數估計法評估之研究". Thesis, 2002. http://ndltd.ncl.edu.tw/handle/07523232327638567936.
Texto completo國防大學中正理工學院
兵器系統工程研究所
90
In the field of reliability engineering, it is very important to accurately estimate the parameters of system (or equipment) reliability. System reliability is a critical factor in the frequency of maintenance, and the maintenance frequency has a pivotal impact on logistics support requirements. System operation is highly dependent on the accuracy of prediction of “what” and “when” that an item will failed. Based on the prediction and analysis of reliability, an overall optimum logistic support capability can be developed to establish a complete Integrated Logistics Support Plan (ILSP). Due to the drought of logistics information systems and the difficulty of maintenance data collection, the objective of this research is to develop an effective procedure on system reliability estimation by using field censoring failure data. This research focuses on the type II right censored and double censored Weibull distributed failure data. Different estimators of the Maximum Likelihood Estimates (MLE) and Least squares Estimates (LSE) are compared and analyzed by using the numerical simulation models. The method of gray relational analysis is used to sort the performance of several estimators by the simulation outputs including the mean, median, and root mean square error (RMSE) of different Weibull distributions. Traditionally, in comparison of parameter estimators, both scale and shape parameters of Weibull distribution are considered and evaluated together. For Weibull distributed type II censoring data, the results of this research suggest that using MLE estimator to estimate scale parameters and LSE to shape parameters can derive not only a more accurate set of parameters, but more accurate reliability estimations.
CHEN, GUO-LIANG, та 陳國良. "逐步加壓加速壽命測試下的參數估計". Thesis, 1990. http://ndltd.ncl.edu.tw/handle/23643604080744993595.
Texto completoHuang, LieShuang, та 黃麗霜. "季節性長距相關模式之參數估計比較". Thesis, 2001. http://ndltd.ncl.edu.tw/handle/08497005629798262354.
Texto completoLIN, JI-PING, та 林季平. "對數線性模式最大概似參數估計值方向搜尋法". Thesis, 1990. http://ndltd.ncl.edu.tw/handle/25913081606917435367.
Texto completohuang, I.-Wen, та 黃怡紋. "偽概似函數在馬可夫隨機域中模型參數的估計". Thesis, 2004. http://ndltd.ncl.edu.tw/handle/41756330624674681176.
Texto completoKuo, Jane-Chen, та 郭建成. "實驗計劃法應用於產品可靠度參數估計之研究". Thesis, 2002. http://ndltd.ncl.edu.tw/handle/06525110427834053039.
Texto completo國防大學中正理工學院
兵器系統工程研究所
90
The purpose of this research is to use both Monte Carlo Simulation and Design of Experiment methods to estimate product’s reliability and parameters relative to it. In practice the data collected for estimating are faced with and affected by the following conditions: incomplete data, number of test specimens, right and/or left censoring, etc. However, we are looking forward to understanding the relationships among the different factors and the estimated parameters before making any decision for reliability analysis. This research focuses on two failure modes: Weibull and log-normal distributions. Each distribution’s parameters are assigned and combined with the mentioned factors in different ways. The Monte Carlo Simulation method is used to make up the “true” data for analyzing. On the other hand the Design of Experiment method is used to cut down the number of the factors’ combination. That also means the saving of cost and time. The results of this research are to get the characteristics of reliability for a product and to realize how the effects of the factors on the estimated reliability parameters. The results also show that a “CFR failure rate product” is easier to estimate its reliability parameters. On the contrary a “DFR failure rate product” is worse to do so. In this situation, to increase the number of sampling will get a better result than to decrease the right-left censored time.
李彥緯. "隱藏馬可夫模型之參數估計及模型選擇". Thesis, 1999. http://ndltd.ncl.edu.tw/handle/78747405957231736310.
Texto completo陳嘯虎. "型II設限下Gamma分配之參數估計及其應用". Thesis, 1994. http://ndltd.ncl.edu.tw/handle/65309543323532448798.
Texto completo林雨農. "一階格點資料之邊際效應探討及參數估計". Thesis, 2002. http://ndltd.ncl.edu.tw/handle/82459460042781399559.
Texto completoCHEN, FANG-ZI, та 陳放子. "位置參數的強韌L-估計在小樣本中的研究". Thesis, 1986. http://ndltd.ncl.edu.tw/handle/22089608285833780627.
Texto completoJIANG, BO-XIANG, та 江柏香. "對混合幾何分配的預測與參數之無母數經驗貝式估計". Thesis, 1990. http://ndltd.ncl.edu.tw/handle/28834824095405772164.
Texto completo謝育泰. "在自變數有測量誤差的情形下模型參數的估計與修正". Thesis, 2001. http://ndltd.ncl.edu.tw/handle/76797382898126979503.
Texto completo林玠毅. "BILOG-MG與NOHARM在二向度測驗試題參數估計之比較". Thesis, 2008. http://ndltd.ncl.edu.tw/handle/67333341945298750101.
Texto completo國立彰化師範大學
統計資訊研究所
96
Unidimensionality is one of the most important assumptions in item response theory (IRT). However, many researchers find that educational tests are often sensitive to multiple abilities. The use of IRT-based computer programs to analyze non-unidimensional tests seems inappropriate. Take English test of TASA in Taiwan for example, although the test consists of listening comprehension and reading, TASA utilizes the unidimensional software, BILOG-MG, to analyze test data. The object of this research is to investigate the performance of BILOG-MG as long as NOHARM in estimating multidimensional item parameters MDISC and MDIFF. Several tests with different two-dimensional item structures are simulated. Bias and RMSE of each estimate are present and carefully discussed in the study. The results show that NOHARM obtains more accurate estimates than BILOG-MG does.
Liu, Chiao-Ling, та 劉巧鈴. "單因子Heath-Jarrow-Morton模型隱含參數之估計~T-Bill分析". Thesis, 2001. http://ndltd.ncl.edu.tw/handle/87137434631209382996.
Texto completo元智大學
管理研究所
89
ABSTRCT Our main motivation is to estimate implied parameters of the HJM model and test the effectiveness of their model. The estimating problem associated with the HJM model is that there is a stochastic term in the forward bond price formula. Musiela, Turnbull and Wakeman (1993) and the following study by Jarrow and Turnbull (1994) derive closed-form solutions for bonds and their associated derivatives. Therefore, it is easier to estimate the parameters for us. In order to estimate the parameters implied HJM model, we use the bond price formula with the closed-form solutions to conduct the empirical systems .Our estimation technique uses the generalized method of moments (GMM). To test the effectiveness of their model ，the standard deviation of the returns of the portfolio is used as the measure of hedging effectiveness. The lower the standard deviation, the better is the performance. The hedging effectiveness between delta hedge and modified Macaulay duration is compared. Key word：HJM，GMM，Hedge
Su, Tsui-Mei, та 蘇翠湄. "加法風險模式之參數估計及其區域混淆現象". Thesis, 2001. http://ndltd.ncl.edu.tw/handle/33636528069429778959.
Texto completoZHANG, YOU-FU, та 張右復. "時間序列模型中缺失資料下參數估計之研究". Thesis, 1987. http://ndltd.ncl.edu.tw/handle/76572518389919150075.
Texto completoYANG, JIN-ZHANG, та 楊錦章. "在加權損失函數之下多個負二項式分配參數的同時估計". Thesis, 1988. http://ndltd.ncl.edu.tw/handle/86051715968764609639.
Texto completoLee, Hui-Wen, та 李惠文. "二分有測量錯誤的連續自變數對邏輯迴歸參數估計的影響". Thesis, 2000. http://ndltd.ncl.edu.tw/handle/46135257242607653341.
Texto completoHung, Wen-Hsueh, та 洪文. "遺傳演算法於降雨逕流模式參數最佳估計之研究". Thesis, 2002. http://ndltd.ncl.edu.tw/handle/95872386524240346181.
Texto completo國立臺灣大學
土木工程學研究所
90
Genetic Algorithms, GAs, were first developed in the 1960s by John Holland to explain and model the adaptation of natural systems. Computationally simple and powerful, GA is well known to be fit for multi-objective optimization problems. In this research, a rainfall-runoff model of Wu-Creek basin is constructed based on the Tank Model theory to make sure that all hydrologic properties are coincident. Then a GA-based method, MOGA, is used to estimate parameters in Tank Model. During optimization process, three objective functions are considered, and two typhoon events, Amber and Zeb, are used for the calibration of the model parameters. The simulated results are in good agreement with the observed. It shows that genetic algorithm is quite capable of solving multi-objective optimization problems.
Cheng, Pi-Wen, та 鄭弼文. "基於廣義隱藏式馬可夫模型之參數估計蒙地卡羅模擬研究基於廣義隱藏式馬可夫模型之參數估計蒙地卡羅模擬研究基於廣義隱藏式馬可夫模型之參數估計蒙地卡羅模擬研究". Thesis, 2006. http://ndltd.ncl.edu.tw/handle/49129451022161216752.
Texto completo國立臺中教育大學
教育測驗統計研究所
94
The symbol probability matrix at every time point must be the same in the original Hidden Markov Model（HMM）. HMM is treated as “Fixed Hidden Markov Model” and is only applied for the testing analysis which has the same symbol probability. Hence, the suitable scope of HMM is quite limited. This study expends the definition of “Hidden Markov Model” by including dynamic symbol probability. We call the expended definition as “Generalizing Hidden Markov Model; GHMM”. The study applies proper parameter estimation algorithm to estimate so that the usability of GHMM is not only for on-line-testing with transiting probability but also for general paper-pencil-testing with item random answering. A theoretical framework of GHMM model is introduced in the study. Furthermore, we design a usable program for parameter estimation of the testing model based on GHMM model in a Monte Carlo simulation study, and make sure that the above model is useful by testing in the elementary school. Finally, we propose the best timing to use GHMM and provide suggestions for the teaching researchers.
Jie, Li Shian, та 李憲杰. "一般化自迴歸條件異質性變異數模型參數之選定、估計與檢定". Thesis, 1994. http://ndltd.ncl.edu.tw/handle/38933872361394211753.
Texto completo國立成功大學
工業管理研究所
82
Since the introduction by Bollerslev in 1986, GARCH models have been broadly applied to financial analysis. However, due to the lack of a systematic method on how to determine the ap- propriate numbers of the model parameters, researchers usually use their subjective judgments and set p=1 and q=1. In this research, we introduce a systematic approach to resolve the above problem. More specifically, we recommend using AIC, SIC, and FIC criteria to select models and using maximum likelihood method to estimate and test parameters. Finally, we present a new criterion to evaluate model performance. To illustrate this selection process, we use Taiwan''''s daily weighted stock index as an example. Our analysis indicates that the conditional heteroscedasticity does exist in the returns of the stock index . We also fit the data to several ARMA and GARCH models and use one-day-ahead forecast to come up with the prediction intervals and cumulate absolute errors. We conclude that the GARCH(1,1) model can be used to explain the inherent volatility behavior of the returns of the stock index.
林慧清. "多元計分三參數試題選項分析模式半最大概似估計法之研究". Thesis, 2002. http://ndltd.ncl.edu.tw/handle/90370992347068292730.
Texto completo臺中師範學院
教育測驗統計研究所
90
In order to analysis multiple choice response data, it should be used multiple choice items model appropriately, but nowadays, all the theories used by test is bivariate logistic model. In accordance with this shortcoming, Hsiang-Chuan Liu(2000) published 「Multivariate logistic response fixed effect model for multiple choice items」 ,indicated unequivocally that this model not only give the ICC of distractor options but also consider the effect of distractor power of distractor options are not necessarily same, so it strengthen the sufficient, effectiveness of correct options ICC. In another word, the information of parameter of a test by multiple choice items model is more sufficient than by parameter bivariate logistic model. We make a study of the model in this research. We consider that the estimative trouble of ability, the restriction of sample size on establishing model, and study the estimative method and the accuracy of parameter estimation. In this model we estimate ability and item parameters by different methods. The estimation method of ability is 「Kernel smoothing approaches to nonparametric item characteristic curve estimation with a weight」. First, we can estimate ability parameter directly, then estimate the item option characteristic parameters by maximum likelihood estimation (MLE), we call this method semi-MLE.
王嘉煒. "以核估計量建構點過程強度函數之信賴區帶其平滑參數之探討". Thesis, 2003. http://ndltd.ncl.edu.tw/handle/53003204969606518278.
Texto completoWEI, HUI-O., та 魏輝娥. "最適尾端參數估計之探討：台灣股票報酬風險值之應用". Thesis, 2003. http://ndltd.ncl.edu.tw/handle/36420742785609252772.
Texto completo國立中正大學
國際經濟研究所
91
As a measure of the exposed market risk for a financial asset, Value at Risk (VaR) has become the most popular method. Among numerous calculations of VaRs, methods based on extreme value theory (EVT) have been studied considerably in theorectical and empirical literature. Peak over threshold (POT) method and the block maximum (or minimum) method (BMM) are two alternatives for implemeting EVT. The difficulty of determining the threshold value in POT and the block size in BMM makes these two methods problem some. It has well been explored that the estimate of tail index using Hill estimator is sensitive to the threshold value. In this thesis, an optimal (in the sense of mean squared errors being minimized) threshold determined by the method suggested by Danielsson et al. (1997a) is investigated. As the mean squared error is calculated via two-step sub-sample bootstrap method, the effects of various bootstrap procedures on the determination of optimal threshold are particularly studied in this thesis. Typically, the bootstrapping methods designed for time series data are considered, such as the moving block method and stationary bootstrapping method. Based on our empirical studies on Taiwan’s aggregate and eight industrial stock indexes, VaRs calculated with different optimal thresholds determined by bootstrapping procedures outperform the ones with the threshold determined by the 5% principle of Tsay (2001) in accuracy, especially for the more extreme VaRs. Among the bootstrapping methods considered in this thesis, our empirical results confirm the superiority moving block procedure.
Yi-Ying, Lin, та 林依穎. "BILOG-MG與SCORIGHT在NEAT設計不同測驗組合下能力及試題參數估計效果之比較". Thesis, 2010. http://ndltd.ncl.edu.tw/handle/02612438331668262244.
Texto completo國立臺中教育大學
教育測驗統計研究所
99
This research is to estimate the root mean square error (RMSE) of ability parameters and item parameters by BILOG-MG and SCORIGHT when the ability distribution is normal or bimodal in non-equivalent groups with anchor test design(NEAT). Four independent factors are considered in this research：(1)the sample sizes；(2)the numbers of examinees；(3) the proportion of the testlet items；(4) the variances of the testlet effects . The results of this research show that: 1. The RMSE of the ability and item difficulty parameter decreases as the sample size or increases the number of examinees when the ability distribution is normal. 2. The RMSE of the ability and item difficulty parameter decreases as the sample size decreases or the number of examinees and the proportion of the testlet item increase when the ability distribution is bimodal. 3. When the ability distribution is normal, compare with BILOG-MG and SCORIGHT simulation, the RMSE of the item discrimination and difficulty parameters are smaller under SCORIGHT simulation. 4. When the ability distribution is bimodal, compare with BILOG-MG and SCORIGHT simulation, the RMSE of the ability and the item parameters are smaller under BILOG-MG simulation.
DDHE, RONG-GUI, та 何榮桂. "題庫中項目參數分配型態對電腦化適性測驗能力估計的影響". Thesis, 1990. http://ndltd.ncl.edu.tw/handle/97317398186524631562.
Texto completo謝雅青. "失業勞工數位學習成效評估之研究—以輔助參加提升數位能力研習計畫者為對象". Thesis, 2007. http://ndltd.ncl.edu.tw/handle/85269011500047168885.
Texto completo