Global ETD Search

181	Inférence robuste à la présence des valeurs aberrantes dans les enquêtes Dongmo Jiongo, Valéry 12 1900 (has links) Cette thèse comporte trois articles dont un est publié et deux en préparation. Le sujet central de la thèse porte sur le traitement des valeurs aberrantes représentatives dans deux aspects importants des enquêtes que sont : l’estimation des petits domaines et l’imputation en présence de non-réponse partielle. En ce qui concerne les petits domaines, les estimateurs robustes dans le cadre des modèles au niveau des unités ont été étudiés. Sinha & Rao (2009) proposent une version robuste du meilleur prédicteur linéaire sans biais empirique pour la moyenne des petits domaines. Leur estimateur robuste est de type «plugin», et à la lumière des travaux de Chambers (1986), cet estimateur peut être biaisé dans certaines situations. Chambers et al. (2014) proposent un estimateur corrigé du biais. En outre, un estimateur de l’erreur quadratique moyenne a été associé à ces estimateurs ponctuels. Sinha & Rao (2009) proposent une procédure bootstrap paramétrique pour estimer l’erreur quadratique moyenne. Des méthodes analytiques sont proposées dans Chambers et al. (2014). Cependant, leur validité théorique n’a pas été établie et leurs performances empiriques ne sont pas pleinement satisfaisantes. Ici, nous examinons deux nouvelles approches pour obtenir une version robuste du meilleur prédicteur linéaire sans biais empirique : la première est fondée sur les travaux de Chambers (1986), et la deuxième est basée sur le concept de biais conditionnel comme mesure de l’influence d’une unité de la population. Ces deux classes d’estimateurs robustes des petits domaines incluent également un terme de correction pour le biais. Cependant, ils utilisent tous les deux l’information disponible dans tous les domaines contrairement à celui de Chambers et al. (2014) qui utilise uniquement l’information disponible dans le domaine d’intérêt. Dans certaines situations, un biais non négligeable est possible pour l’estimateur de Sinha & Rao (2009), alors que les estimateurs proposés exhibent un faible biais pour un choix approprié de la fonction d’influence et de la constante de robustesse. Les simulations Monte Carlo sont effectuées, et les comparaisons sont faites entre les estimateurs proposés et ceux de Sinha & Rao (2009) et de Chambers et al. (2014). Les résultats montrent que les estimateurs de Sinha & Rao (2009) et de Chambers et al. (2014) peuvent avoir un biais important, alors que les estimateurs proposés ont une meilleure performance en termes de biais et d’erreur quadratique moyenne. En outre, nous proposons une nouvelle procédure bootstrap pour l’estimation de l’erreur quadratique moyenne des estimateurs robustes des petits domaines. Contrairement aux procédures existantes, nous montrons formellement la validité asymptotique de la méthode bootstrap proposée. Par ailleurs, la méthode proposée est semi-paramétrique, c’est-à-dire, elle n’est pas assujettie à une hypothèse sur les distributions des erreurs ou des effets aléatoires. Ainsi, elle est particulièrement attrayante et plus largement applicable. Nous examinons les performances de notre procédure bootstrap avec les simulations Monte Carlo. Les résultats montrent que notre procédure performe bien et surtout performe mieux que tous les compétiteurs étudiés. Une application de la méthode proposée est illustrée en analysant les données réelles contenant des valeurs aberrantes de Battese, Harter & Fuller (1988). S’agissant de l’imputation en présence de non-réponse partielle, certaines formes d’imputation simple ont été étudiées. L’imputation par la régression déterministe entre les classes, qui inclut l’imputation par le ratio et l’imputation par la moyenne sont souvent utilisées dans les enquêtes. Ces méthodes d’imputation peuvent conduire à des estimateurs imputés biaisés si le modèle d’imputation ou le modèle de non-réponse n’est pas correctement spécifié. Des estimateurs doublement robustes ont été développés dans les années récentes. Ces estimateurs sont sans biais si l’un au moins des modèles d’imputation ou de non-réponse est bien spécifié. Cependant, en présence des valeurs aberrantes, les estimateurs imputés doublement robustes peuvent être très instables. En utilisant le concept de biais conditionnel, nous proposons une version robuste aux valeurs aberrantes de l’estimateur doublement robuste. Les résultats des études par simulations montrent que l’estimateur proposé performe bien pour un choix approprié de la constante de robustesse. / This thesis focuses on the treatment of representative outliers in two important aspects of surveys: small area estimation and imputation for item non-response. Concerning small area estimation, robust estimators in unit-level models have been studied. Sinha & Rao (2009) proposed estimation procedures designed for small area means, based on robustified maximum likelihood parameters estimates of linear mixed model and robust empirical best linear unbiased predictors of the random effect of the underlying model. Their robust methods for estimating area means are of the plug-in type, and in view of the results of Chambers (1986), the resulting robust estimators may be biased in some situations. Biascorrected estimators have been proposed by Chambers et al. (2014). In addition, these robust small area estimators were associated with the estimation of the Mean Square Error (MSE). Sinha & Rao (2009) proposed a parametric bootstrap procedure based on the robust estimates of the parameters of the underlying linear mixed model to estimate the MSE. Analytical procedures for the estimation of the MSE have been proposed in Chambers et al. (2014). However, their theoretical validity has not been formally established and their empirical performances are not fully satisfactorily. Here, we investigate two new approaches for the robust version the best empirical unbiased estimator: the first one relies on the work of Chambers (1986), while the second proposal uses the concept of conditional bias as an influence measure to assess the impact of units in the population. These two classes of robust small area estimators also include a correction term for the bias. However, they are both fully bias-corrected, in the sense that the correction term takes into account the potential impact of the other domains on the small area of interest unlike the one of Chambers et al. (2014) which focuses only on the domain of interest. Under certain conditions, non-negligible bias is expected for the Sinha-Rao method, while the proposed methods exhibit significant bias reduction, controlled by appropriate choices of the influence function and tuning constants. Monte Carlo simulations are conducted, and comparisons are made between: the new robust estimators, the Sinha-Rao estimator, and the bias-corrected estimator. Empirical results suggest that the Sinha-Rao method and the bias-adjusted estimator of Chambers et al (2014) may exhibit a large bias, while the new procedures offer often better performances in terms of bias and mean squared error. In addition, we propose a new bootstrap procedure for MSE estimation of robust small area predictors. Unlike existing approaches, we formally prove the asymptotic validity of the proposed bootstrap method. Moreover, the proposed method is semi-parametric, i.e., it does not rely on specific distributional assumptions about the errors and random effects of the unit-level model underlying the small-area estimation, thus it is particularly attractive and more widely applicable. We assess the finite sample performance of our bootstrap estimator through Monte Carlo simulations. The results show that our procedure performs satisfactorily well and outperforms existing ones. Application of the proposed method is illustrated by analyzing a well-known outlier-contaminated small county crops area data from North-Central Iowa farms and Landsat satellite images. Concerning imputation in the presence of item non-response, some single imputation methods have been studied. The deterministic regression imputation, which includes the ratio imputation and mean imputation are often used in surveys. These imputation methods may lead to biased imputed estimators if the imputation model or the non-response model is not properly specified. Recently, doubly robust imputed estimators have been developed. However, in the presence of outliers, the doubly robust imputed estimators can be very unstable. Using the concept of conditional bias as a measure of influence (Beaumont, Haziza and Ruiz-Gazen, 2013), we propose an outlier robust version of the doubly robust imputed estimator. Thus this estimator is denoted as a triple robust imputed estimator. The results of simulation studies show that the proposed estimator performs satisfactorily well for an appropriate choice of the tuning constant. Estimateur corrigé pour le biais Biais conditionnel Valeurs aberrantes Inférence basée sur le modèle Inférence basée sur le plan Petits domaines Bootstrap Modèle linéaire mixte Robustesse Imputation Corrected-bias estimator Conditional bias Outliers Model-based inference Sampling-based inference Small-area Linear mixed model Robustness
182	變數轉換之穩健迴歸分析張嘉璁 Unknown Date (has links) 在傳統的線性迴歸分析當中，當基本假設不滿足時，有時可考慮變數轉換使得資料能夠比較符合基本假設。在眾多的轉換方法當中，以Box和Cox(1964)所提出的乘冪轉換(Box-Cox power transformation)最為常用，乘冪轉換可將某些複雜的系統轉換成線性常態模式。然而當資料存在離群值(outlier)時，Box-Cox Transformation會受到影響，因此不是一種穩健方法。在本篇論文當中，我們利用前進演算法(forward search algorithm)求得最小消去平方估計量(Least trimmed squares estimator)，在過程當中估計出穩健的轉換參數。 Box-Cox乘冪轉換前進演算法最小中位數平方法最小消去平方法分數統計量穩健迴歸損壞點離群值 Box-Cox power transformation Forward search algorithm Least median of squares Least trimmed squares Score statistic Robust regression Breakdown point Outliers
183	Sélection de modèles robuste : régression linéaire et algorithme à sauts réversibles Gagnon, Philippe 10 1900 (has links) No description available. analyse en composantes principales inférence bayésienne robustesse valeurs aberrantes Bayesian inference Markov chain Monte Carlo methods Outliers Principal component analysis Random walk Metropolis algorithm Robustness Super heavy-tailed distributions
184	Seguro contra risco de downside de uma carteira: uma proposta híbrida frequentista-Bayesiana com uso de derivativos Pérgola, Gabriel Campos 23 January 2013 (has links) Submitted by Gabriel Campos Pérgola (gabrielpergola@gmail.com) on 2013-02-04T12:56:43Z No. of bitstreams: 1 DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) / Rejected by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br), reason: Prezado Gabriel, Não recebemos os arquivo em PDF. Att. Suzi 3799-7876 on 2013-02-05T18:53:00Z (GMT) / Submitted by Gabriel Campos Pérgola (gabrielpergola@gmail.com) on 2013-02-05T19:00:17Z No. of bitstreams: 2 DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2013-02-05T19:07:12Z (GMT) No. of bitstreams: 2 DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) / Made available in DSpace on 2013-02-05T19:09:04Z (GMT). No. of bitstreams: 2 DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) DissertationGabrielPergola2013.pdf: 521205 bytes, checksum: 85369078a82b0d5cc02f8248961e9214 (MD5) Previous issue date: 23-01-13 / Portfolio insurance allows a manager to limit downside risk while allowing participation in upside markets. The purpose of this dissertation is to introduce a framework to portfolio insurance optimization from a hybrid frequentist-Bayesian approach. We obtain the joint distribution of regular returns from a frequentist statistical method, once the outliers have been identified and removed from the data sample. The joint distribution of extreme returns, in its turn, is modelled by a Bayesian network, whose topology reflects the events that can significantly impact the portfolio performance. Once we link the regular and extreme distributions of returns, we simulate future scenarios for the portfolio value. The insurance subportfolio is then optimized by the Differential Evolution algorithm. We show the framework in a step by step example for a long portfolio including stocks participating in the Bovespa Index (Ibovespa), using market data from 2008 to 2012. / Seguros de carteiras proporcionam aos gestores limitar o risco de downside sem renunciar a movimentos de upside. Nesta dissertação, propomos um arcabouço de otimização de seguro de carteira a partir de um modelo híbrido frequentista-Bayesiano com uso de derivativos. Obtemos a distribuição conjunta de retornos regulares através de uma abordagem estatística frequentista, uma vez removidos os outliers da amostra. A distribuição conjunta dos retornos extremos, por sua vez, é modelada através de Redes Bayesianas, cuja topologia contempla os eventos que o gestor considera crítico ao desempenho da carteira. Unindo as distribuições de retornos regulares e extremos, simulamos cenários futuros para a carteira. O seguro é, então, otimizado através do algoritmo Evolução Diferencial. Mostramos uma aplicação passo a passo para uma carteira comprada em ações do Ibovespa, utilizando dados de mercado entre 2008 e 2012. Seguro de carteira Modelos híbridos Modelos multivariados de retornos Identificação de outliers Minimum Covariance Determinant Distribuição Hiperbólica Generalizada Redes bayesianas Simulação Otimização Algoritmo Evolução Diferencial Portfolio insurance Hybrid methods Generalized hyperbolic distribution Bayesian nets Simulation Economia Derivativos (Finanças) Algoritmos
185	異常住宅價格檢測與處理之研究－以個別估價觀點分析 / The study of singular residential price detection and management - with the valuations by appraisers' perspective 高裕政, Kao, Yu Cheng Unknown Date (has links) 國內近年來有許多文獻在進行特徵價格模型預測時，避免樣本中存在異常點會造成模型估計值產生偏差，會使用統計軟體進行異常點檢測，但皆是直接將檢測出的異常點刪除，未加以著墨探究這些異常點的特徵結構、成因及特色等。因此，本研究透過統計檢定方法，探討刪除異常點前後整體樣本的特徵結構變化，並以個別估價觀點加以探討住宅交易樣本異常點的成因與特色，藉此歸納出實價登錄資料未揭露的重要特徵，以及迴歸模型搜尋疑似申報不實案件之可行性。透過敘述統計及樣本結構差異檢定結果發現，異常樣本的離散程度相對原始樣本與正常樣本較大，且經過刪除異常點的正常樣本特徵結構差異程度縮小；異常點的形成可能受到區位變數無法反映實際情況及樣本群聚程度影響，也可能因模型未納入某些重要的特徵變數，而使隱含該變數的樣本被判斷為異常點；異常樣本與正常樣本的成交總價、土地坪數、建物坪數、總樓層、所在樓層及屋齡等變數平均數、變異數及中位數有顯著差異。藉由個案分析結果歸納，可能因異常個案的住宅屬性存在整幢大樓住商混合使用、特殊鄰居、附屬建物占比過高、高總價豪宅產品、都更效益、增建效益、裝潢效益、約定專用空間效益、樓層高度挑高、獨特視野景觀或特殊區位條件；外部環境存在鄰近嫌惡設施或迎毗設施；交易情況存在買方身分特殊之影響，但受限於實價登錄未要求登載並揭露這些特徵，故模型未考量這些因素對價格的影響，使得模型可能將隱含這些特徵的樣本判斷為異常點，並進而影響模型預測結果。另外也發現，實價登錄資料存在登載錯誤及價格申報不實的情況，且可能被模型判斷為異常點。 / Many literatures use statistics-way to detect outliers in preventing any extreme deviation in hedonic price model prediction. Nevertheless, deleting the outliers instead of investigation into the structures, causes and features. Hence, this thesis studies the feature structures variation of the sample before and after deleting the outliers and with the valuations by appraisers’ perspective to inquire into the factors and features of the outliers in residential transactions. Thereby to summarize the significant features that are not disclosed by real price registration and feasibility in searching the possible false declaration of price by regression. Through descriptive statistics and sample structural difference parametric and nonparametric test shows the discreteness level of singular (outliers only) samples is greater than the primary (outliers including) and normal (outliers deleting) samples and the feature structure variation lessened after deleting the outliers in normal samples. The formation of outliers may be influenced by location variable not able to reflect actual circumstances and level of clustering in samples. Maybe some significant variables are not subsumed into the model, which leads to the judgement of samples with this variable to be outliers. The mean, variance and median in total traded price, land size, building size, total floors, exact floor and house age of singular samples are notably different with normal ones. With the analysis of cases, the possible reasons may be residential and commercial mixed-use in building, peculiar neighbors, high proportion of accessory building, luxury houses, urban renewal benefits, building addition benefits, interior decoration benefits, agreed space benefits, high-ceiling benefits, unique view or location, YIMBY and NIMBY property in environment and special relationship between the buyer and seller. Nevertheless, due to the nondisclosure of these features in real price registration that the model does not take these into consideration. That leads to the judgement of samples with these features as outliers and affects the model prediction. Also the registration error and false declaration in price may also be judged as outliers. 異常點特徵價格模型個別估價實價登錄不動產資訊透明住宅價格申報不實 Outliers Hedonic price model Appraiser Real price registration Transparency of real estate information False declaration of house price
186	AUTOMATED OPTIMAL FORECASTING OF UNIVARIATE MONITORING PROCESSES : Employing a novel optimal forecast methodology to define four classes of forecast approaches and testing them on real-life monitoring processes Razroev, Stanislav January 2019 (has links) This work aims to explore practical one-step-ahead forecasting of structurally changing data, an unstable behaviour, that real-life data connected to human activity often exhibit. This setting can be characterized as monitoring process. Various forecast models, methods and approaches can range from being simple and computationally "cheap" to very sophisticated and computationally "expensive". Moreover, different forecast methods handle different data-patterns and structural changes differently: for some particular data types or data intervals some particular forecast methods are better than the others, something that is usually not known beforehand. This raises a question: "Can one design a forecast procedure, that effectively and optimally switches between various forecast methods, adapting the forecast methods usage to the changes in the incoming data flow?" The thesis answers this question by introducing optimality concept, that allows optimal switching between simultaneously executed forecast methods, thus "tailoring" forecast methods to the changes in the data. It is also shown, how another forecast approach: combinational forecasting, where forecast methods are combined using weighted average, can be utilized by optimality principle and can therefore benefit from it. Thus, four classes of forecast results can be considered and compared: basic forecast methods, basic optimality, combinational forecasting, and combinational optimality. The thesis shows, that the usage of optimality gives results, where most of the time optimality is no worse or better than the best of forecast methods, that optimality is based on. Optimality reduces also scattering from multitude of various forecast suggestions to a single number or only a few numbers (in a controllable fashion). Optimality gives additionally lower bound for optimal forecasting: the hypothetically best achievable forecast result. The main conclusion is that optimality approach makes more or less obsolete other traditional ways of treating the monitoring processes: trying to find the single best forecast method for some structurally changing data. This search still can be sought, of course, but it is best done within optimality approach as its innate component. All this makes the proposed optimality approach for forecasting purposes a valid "representative" of a more broad ensemble approach (which likewise motivated development of now popular Ensemble Learning concept as a valid part of Machine Learning framework). / Denna avhandling syftar till undersöka en praktisk ett-steg-i-taget prediktering av strukturmässigt skiftande data, ett icke-stabilt beteende som verkliga data kopplade till människoaktiviteter ofta demonstrerar. Denna uppsättning kan alltså karakteriseras som övervakningsprocess eller monitoringsprocess. Olika prediktionsmodeller, metoder och tillvägagångssätt kan variera från att vara enkla och "beräkningsbilliga" till sofistikerade och "beräkningsdyra". Olika prediktionsmetoder hanterar dessutom olika mönster eller strukturförändringar i data på olika sätt: för vissa typer av data eller vissa dataintervall är vissa prediktionsmetoder bättre än andra, vilket inte brukar vara känt i förväg. Detta väcker en fråga: "Kan man skapa en predictionsprocedur, som effektivt och på ett optimalt sätt skulle byta mellan olika prediktionsmetoder och för att adaptera dess användning till ändringar i inkommande dataflöde?" Avhandlingen svarar på frågan genom att introducera optimalitetskoncept eller optimalitet, något som tillåter ett optimalbyte mellan parallellt utförda prediktionsmetoder, för att på så sätt skräddarsy prediktionsmetoder till förändringar i data. Det visas också, hur ett annat prediktionstillvägagångssätt: kombinationsprediktering, där olika prediktionsmetoder kombineras med hjälp av viktat medelvärde, kan utnyttjas av optimalitetsprincipen och därmed få nytta av den. Alltså, fyra klasser av prediktionsresultat kan betraktas och jämföras: basprediktionsmetoder, basoptimalitet, kombinationsprediktering och kombinationsoptimalitet. Denna avhandling visar, att användning av optimalitet ger resultat, där optimaliteten för det mesta inte är sämre eller bättre än den bästa av enskilda prediktionsmetoder, som själva optimaliteten är baserad på. Optimalitet reducerar också spridningen från mängden av olika prediktionsförslag till ett tal eller bara några enstaka tal (på ett kontrollerat sätt). Optimalitet producerar ytterligare en nedre gräns för optimalprediktion: det hypotetiskt bästa uppnåeliga prediktionsresultatet. Huvudslutsatsen är följande: optimalitetstillvägagångssätt gör att andra traditionella sätt att ta hand om övervakningsprocesser blir mer eller mindre föråldrade: att leta bara efter den enda bästa enskilda prediktionsmetoden för data med strukturskift. Sådan sökning kan fortfarande göras, men det är bäst att göra den inom optimalitetstillvägagångssättet, där den ingår som en naturlig komponent. Allt detta gör det föreslagna optimalitetstillvägagångssättetet för prediktionsändamål till en giltig "representant" för det mer allmäna ensembletillvägagångssättet (något som också motiverade utvecklingen av numera populär Ensembleinlärning som en giltig del av Maskininlärning). predictions forecasting optimal forecasting forecast classes optimality rules ensemble forecasting state switching combinational forecasting optimality framework exponential smoothing ARIMA ARMA SARIMA Double-Seasonal Holt-Winters time series Wikipedia Wikimedia Wikipedia data Twitter Twitter data electricity data monitoring processes monitoring process monitoring error metrics outliers missing values Mathematics Matematik
187	Statistical Machine Learning in Biomedical Engineering González Cebrián, Alba 15 April 2024 (has links) [ES] Esta tesis, desarrollada bajo una beca de formación de personal investigador de la Universitat Politècnica de València, tiene como objetivo proponer y aplicar metodologías de Statistical Machine Learning en contextos de Ingeniería Biomédica. Este concepto pretende aunar el uso de modelos de aprendizaje automático junto con la búsqueda de comprensión e interpretabilidad clásica del razonamiento estadístico, dando lugar a soluciones tecnológicas de problemas biomédicos que no pasen únicamente por el objetivo de optimizar el desempeño predictivo de los modelos. Para ello, se han dibujado dos objetivos principales que vertebran además el documento: proponer metodologías novedosas dentro del paraguas del Statistical Machine Learning, y aplicar soluciones a problemas biomédicos reales manteniendo esta filosofía en mente. Estos objetivos se han materializado en contribuciones metodológicas para la simulación de valores atípicos y la imputación de datos faltantes en presencia de datos atípicos, y en contribuciones aplicadas a casos reales para la mejora de procesos de atención médica, la mejora en el diagnóstico y pronóstico de enfermedades, y la estandarización de procedimientos de medición en entornos biotecnológicos. Dichas contribuciones se han artículado en capítulos correspondientes a las dos partes principales ya mencionadas. Finalmente, las conclusiones y líneas futuras cierran el documento, recalcando los mensajes principales de las contribuciones de la tesis doctoral en general, y sentando además las bases para líneas futuras que se han dibujado a consecuencia del trabajo realizado a lo largo del doctorado. / [CA] Aquesta tesi, desenvolupada sota una beca de formació de personal investigador de la Universitat Politècnica de València, té com a objectiu proposar i aplicar metodologies de Statistical Machine Learning en contextos d'Enginyeria Biomèdica. Aquest concepte pretén unir l'ús de models d'aprenentatge automàtic juntament amb la cerca de comprensió i interpretació clàssica del raonament estadístic, donant lloc a solucions tecnològiques de problemes biomèdics que no passen únicament per l'objectiu d'optimitzar el rendiment predictiu dels models. Per a això, s'han dibuixat dos objectius principals que vertebren a més el document: proposar metodologies noves dins del paraigua del Statistical Machine Learning, i aplicar solucions a problemes biomèdics reals mantenint aquesta filosofia en ment. Aquests objectius s'han materialitzat en contribucions metodològiques per a la simulació de valors atípics i la imputació de dades mancants en presència de valors atípics, i en contribucions aplicades a casos reals per a la millora de processos d'atenció mèdica, la millora en el diagnòstic i pronòstic de malalties, i l'estandardització de procediments de mesurament en entorns biotecnològics. Aquestes contribucions s'han articulat en capítols corresponents a les dues parts principals ja esmentades. Finalment, les conclusions i línies futures tanquen el document, recalant els missatges principals de les contribucions, de la tesi doctoral en general, i assentant a més les bases per a línies futures que s'han dibuixat com a consequència del treball realitzat al llarg del doctorat. / [EN] This thesis, developed under a research personnel formation grant from the Universitat Politècnica de València, aims to propose and apply methodologies of Statistical Machine Learning in Biomedical Engineering contexts. This concept seeks to combine machine learning models with the classic understanding and interpretability of statistical reasoning, resulting in technological solutions for biomedical problems that go beyond solely optimizing the predictive performance of models. To achieve this, two main objectives have been outlined, which also structure the document: proposing novel methodologies within the umbrella of Statistical Machine Learning and applying solutions to real biomedical problems while keeping this philosophy in mind. These objectives have materialized into methodological contributions for simulating outliers and imputing missing data in the presence of outliers and applied contributions to real cases for improving healthcare processes, enhancing disease diagnosis and prognosis, and standardizing measurement procedures in biotechnological environments. These contributions are articulated in chapters corresponding to the aforementioned two main parts. Finally, the conclusions and future lines of research conclude the document, reiterating the main messages of the contributions and the overall doctoral thesis and laying the groundwork for future lines of inquiry stemming from the work conducted throughout the doctorate. / González Cebrián, A. (2024). Statistical Machine Learning in Biomedical Engineering [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/203529 Biomedical engineering Statistical Machine Learning Latent Variable-based Models Multivariate Analysis Principal Component Analysis (PCA) Data Science Outliers Missing Data Medicine 4.0 Modelos de Variables Latentes Medicina 4.0 Ciencia de datos Datos anómalos Datos faltantes Análisis multivariante Ingeniería biomédica ESTADISTICA E INVESTIGACION OPERATIVA
188	Nestandardní úlohy v odstranění rozmazání obrazu / Image Deblurring in Demanding Conditions Kotera, Jan January 2020 (has links) Title: Image Deblurring in Demanding Conditions Author: Jan Kotera Department: Institute of Information Theory and Automation, Czech Academy of Sciences Supervisor: Doc. Ing. Filip Šroubek, Ph.D., DSc., Institute of Information Theory and Automation, Czech Academy of Sciences Abstract: Image deblurring is a computer vision task consisting of removing blur from image, the objective is to recover the sharp image corresponding to the blurred input. If the nature and shape of the blur is unknown and must be estimated from the input image, image deblurring is called blind and naturally presents a more difficult problem. This thesis focuses on two primary topics related to blind image deblurring. In the first part we work with the standard image deblurring based on the common convolution blur model and present a method of increasing robustness of the deblur- ring to phenomena violating the linear acquisition model, such as for example inten- sity clipping caused by sensor saturation in overexposed pixels. If not properly taken care of, these effects significantly decrease accuracy of the blur estimation and visual quality of the restored image. Rather than tailoring the deblurring method explicitly for each particular type of acquisition model violation we present a general approach based on flexible automatic...
189	GIS-based Episode Reconstruction Using GPS Data for Activity Analysis and Route Choice Modeling / GIS-based Episode Reconstruction Using GPS Data Dalumpines, Ron 26 September 2014 (has links) Most transportation problems arise from individual travel decisions. In response, transportation researchers had been studying individual travel behavior – a growing trend that requires activity data at individual level. Global positioning systems (GPS) and geographical information systems (GIS) have been used to capture and process individual activity data, from determining activity locations to mapping routes to these locations. Potential applications of GPS data seem limitless but our tools and methods to make these data usable lags behind. In response to this need, this dissertation presents a GIS-based toolkit to automatically extract activity episodes from GPS data and derive information related to these episodes from additional data (e.g., road network, land use). The major emphasis of this dissertation is the development of a toolkit for extracting information associated with movements of individuals from GPS data. To be effective, the toolkit has been developed around three design principles: transferability, modularity, and scalability. Two substantive chapters focus on selected components of the toolkit (map-matching, mode detection); another for the entire toolkit. Final substantive chapter demonstrates the toolkit’s potential by comparing route choice models of work and shop trips using inputs generated by the toolkit. There are several tools and methods that capitalize on GPS data, developed within different problem domains. This dissertation contributes to that repository of tools and methods by presenting a suite of tools that can extract all possible information that can be derived from GPS data. Unlike existing tools cited in the transportation literature, the toolkit has been designed to be complete (covers preprocessing up to extracting route attributes), and can work with GPS data alone or in combination with additional data. Moreover, this dissertation contributes to our understanding of route choice decisions for work and shop trips by looking into the combined effects of route attributes and individual characteristics. / Dissertation / Doctor of Philosophy (PhD) GPS time use diary episode extraction multinomial logit travel behavior mode detection episode reconstruction GIS map-matching route choice path size logit potential path area scale estimation Python ArcGIS activity analysis trip reconstruction smartphone global positioning system geographic information system toolkit work trip shop trip potential activity location land use activity episode travel episode stop episode transferability scalability modularity scripting big data travel survey respondent burden preprocessing multipath error tracking purpose detection segmentation data filter data smoothing fuzzy logic neural network decision tree rule-based algorithm trajectory point road network network dataset gateway shortest path horizontal dilution of precision HDOP commonality factor mode transfer point Halifax Nova Scotia Space-Time Activity Research variance inflation factor shapefile comma-separated values traveling salesman problem data mining branch-and-bound algorithm pedestrian network alternative route observed route route efficiency route attributes distance time heading bearing duration acceleration latitude longitude coordinate overlay analysis intersect shopping module data logger walk likelihood ratio test classification table path generation kappa statistic degrees data collection transportation research automate framework ArcToolbox spatio-temporal navigation positioning trace path spatial data topology horizontal accuracy SPSS Stata spatial resolution temporal resolution household survey trip reporting proximity analysis DMTI Desktop Mapping Technologies Inc. road intersection route overlap left turn right turn location analysis time geography spatial statistics buffer analysis cycling bus transit public transportation GEOIDE AGILE data need raw data urban canyon endpoint data cleaning elevation satellite outliers automatic processing nearest node classifier classification method short trip multi-point

Search results