根據研究指出 ETF 資產規模近幾年快速成長，元大台灣卓越 50 基金因市場 規模大等優勢受到投資人的青睞，賴以巨量資料的發展使得文字探勘技術成熟， 故本研究希冀提出一套情感分析的價格預測模型，提升投資者的報酬率。 過往學者以文章中的單詞作為文字探勘的分析單位，常會產生同義詞、多義 詞的問題，因此提出情感型態分析的監督式學習方法建立模型。另外為了解決監 督式學習難以取得訓練資料的限制，本研究混合非監督式學習方法進行主題分群 與情緒傾向標注。 本研究建立台灣股市新聞文本資料集，並篩選熱門議題詞詞庫，進行非監督 式的 LDA 主題模型，發現在 2016 年總統選舉期間，媒體對於公司相關議題的注 意力降低，使得相關的文本數量大幅減少;另外在情緒傾向標注階段，因混和了 NTUSD、知網及自行擴充演算法的情感詞庫，能夠將 10%中性詞彙產生極性判 斷、96%的文本標注情緒傾向。 視覺化工具分析結果指出，DIF-MACD 能夠預測台灣卓越 50 基金的長期走 勢，而新聞情緒指數則在短期的價格波動上表現良好，且在主題模型分群中，總 體經濟、公司維運類別的新聞情緒指數具有約 1-2 日領先指標特性，對於後續的 價格預測模型有所助益。 在監督式情感分析方法，為解決上述同義詞、多義詞的問題，本研究採用型 態分類模型於中文文本，並與向量空間模型、支援向量機等方法做比較。實驗結 果指出優化的型態分類模型，並結合台灣加權股價指數，表現相對良好，F1- Measure 可達 85%。進一步討論新聞情緒對於價格預測的重要性，發現在非交易 時間序列中的新聞情緒，能夠對 0050 的價格波動產生影響。 / The past research points out that the scale of ETF assets has been growing rapidly in recent years. Yuanta Taiwan Top 50 ETF is popular with investors because of the advantages of large market scale. Through the development of Big Data, the technology of Text Mining becomes mature. Thus, we analyze the price forecast model to raise the investors' rate of return. The research of Text Mining used to take the document term to analyze, but it often results in the problem with synonym and polysemy. Therefore, this research proposes a supervised learning method of sentiment pattern analysis. In addition, in order to solve the problem with training data about the supervised learning method, we mix the unsupervised learning method to carry out the subject grouping and sentimental tendency. In this study, we establish the news dataset and screen it as popular terms that are used to an unsupervised method of LDA model. The result points out that the number of news about company dropped significantly during the 2016 Taiwan president election because of the change of media sensation. Moreover, we create the sentiment dictionary that can determine the polarity of 10% neutral terms and the emotional tendency of 96% documents by mixing the NTUSD, HowNet knowledge Database and the self-expansion algorithm. Through the data visualization, the result shows that the curve of DIF-MACD is able to predict the long-term trend of 0050, while the sentiment index of the news makes a good showing in the short-term price volatility. Besides, the news sentiment index of the subjects that belong to general economy and company has about 1 to 2 day leading indicators. Eventually, we employ the Sentiment Pattern Taxonomy Model(PTM) in Chinese texts as supervised learning method and compare with VSM and SVM. The experiment result shows that PTM combined with Taiwan Weighted Stock Index is the best when its F1-Measure is up to 85%. Apart from this, we find that the sentiment index of the news in non-trading time can influence the price volatility of 0050.

Over the past decades, Fourier Transform Infrared (FTIR) microspectroscopy has emerged as a potential candidate to complement Histopathology in the study and diagnosis of tissue diseases. Contrary to the histological examination, which relies on the morphological tissue alterations assessed by visual inspection of stained samples, FTIR chemical imaging is a rapid and label-free tool that provide simultaneously information about histological structures as well as the localisation and magnitude of basic molecular units that compose tissue sections (proteins, nucleic acids, lipids, and carbohydrates). Despite the many proof-of-concept studies demonstrating the effectiveness of FTIR spectroscopy in detecting biological disorders with high levels of sensitivity and specificity, translation into clinical practice has been relatively slow due to the substantial cost of infrared transparent substrates required to collect the images. Thus, the main objective of this research is to evaluate the diagnostic potential of infrared chemical images collected from samples placed on conventional histology glass slides as alternative substrates for FTIR spectroscopy. Swiss mice were submitted to a well-established chemical carcinogenesis protocol, in which cancerous and non-cancerous cutaneous lesions were obtained by varying the exposure time of the animals to carcinogenenic factors. FTIR hyperspectral images were acquired in transmission mode over the mid-infrared region from tissue specimens placed on conventional infrared substrates (calcium fluoride - CaF2) and glass slides. In the first phase of our study, spectral datasets were segmented using k-means (KMCA) and Hierarchical Cluster Analysis (HCA) as clustering algorithms to reconstruct the hyperspectral images aiming to evaluate the ability of the false-color maps in reproducing the histological structures of tissue specimens. The images were segmented by each clustering technique using several different combinations varying parameters including the substrate used to place the samples (CaF2 or conventional glass) and the methods employed to preprocess the datasets. Fingerprint (1000-1800 cm-1) and high wavenumber (3100-4000 cm-1) regions from images collected on CaF2 were separately used as input for image reconstruction and only the high wavenumber range was employed in the case of samples placed on glass. All pseudocolor maps were compared to standard histopathology in order to evaluate the quality and consistency of images after segmentation. KMCA presented slightly superior ability in correctly assigning the pixels of morphochemical maps to the histological structures of the specimen, nevertheless, our findings indicate that the choice of the substrate, input data, preprocessing methods, and sample preparation have more influence in the final results than the clustering algorithm used to reconstruct the images. In the second phase of our study, Principal Component Analysis (PCA) was employed to compare datasets from healthy group to animals exposed to chemicals for 8, 16, and 48 weeks in order to evaluate the biochemical changes induced by chemical carcinogenesis. The performance of classification in each pairwise comparison was calculated using a binary classification test based on Linear Discriminant Analysis associated to PCA (PC-LDA). The method achieved satisfactory discrimination (over 80%) comparing healthy tissue to samples that were classified as papilloma (16 weeks) and invasive squamous cell carcinoma (48 weeks) regardless of the substrate used to place the samples. Statistical measurements obtained comparing healthy skin to animals exposed to carcinogenic factors for 8 weeks (free of malignancy based on the morphological and clinical evidence) ranged from 35-78%, indicating that the ability of PC-LDA in correctly classifying spectral data from cancerous and pre-cancerous lesions vary with the stage of the disease during the tumorigenesis process. Thus, as a proof-of-concept, we demonstrate the feasibility of FTIR spectroscopy in evaluating the biological events triggered by cancer using a label-free methodology that do not rely on expensive substrates and do not disrupt the pathologist workflow. This is a major step forward towards clinical application, since the method can be used to complement the diagnostic process of cancer as a non-subjective alternative that do not require laborious and time-consuming procedures nor expensive probes as biomarkers. / Nas últimas décadas, a microespectroscopia de absorção no infravermelho por transformada de Fourier (FTIR) tem surgido como potencial ferramenta para complementar a Histopatologia no estudo e diagnóstico de doenças teciduais. Ao contrário do exame histológico, que se baseia na inspeção visual de amostras coradas visando avaliar as alterações morfológicas que as doenças ocasionam no tecido, o imageamento químico obtido pela técnica de FTIR baseia-se nas características bioquímicas da amostra sem o uso de colorações. Apesar da vasta literatura comprovando a eficácia da espectroscopia FTIR em detectar alterações biológicas causadas por doenças com altos níveis de sensibilidade e especificidade, a implementação do método na prática clínica tem sido relativamente lenta devido ao alto custo dos substratos transparentes no infravermelho que são necessários para aquisição de dados. Diante disso, o objetivo principal do presente trabalho é avaliar a capacidade diagnóstica de imagens hiperespectrais coletadas de amostras em lâminas de vidro como substratos alternativos para a espectroscopia FTIR. Camundongos Swiss foram submetidos a um protocolo de carcinogênese química, no qual lesões cutâneas cancerosas e não-cancerosas foram obtidas variando-se o tempo de exposição dos animais aos fatores carcinogênicos. Imagens hiperespectrais FTIR foram adquiridas no modo de transmissão na região do infravermelho médio a partir de amostras de tecido depositadas em substratos transparentes no infravermelho (fluoreto de cálcio - CaF2) e vidro convencional. Na primeira fase de nosso estudo, os dados espectrais foram segmentados usando as técnicas estatísticas k-means (KMCA) e Análise Hierárquica de Clusters (HCA) como algoritmos de agrupamento para reconstruir as imagens hiperespectrais com o objetivo de avaliar a capacidade dos mapas de cores falsas em reproduzir as estruturas histológicas das amostras de tecido. As imagens foram segmentadas por cada técnica de agrupamento variando-se o substrato usado para colocar as amostras (CaF2 ou vidro convencional) assim como os métodos de tratamento utilizados para pré-processamento dos dados. As regiões de impressão digital (1000-1800 cm-1) e de altos números de onda (3100-4000 cm-1) das imagens coletadas em CaF2 foram usadas separadamente como dados de entrada para a reconstrução das imagens, enquanto apenas a faixa de altos números de onda foi utilizada no caso de amostras colocadas em vidro. Ao fim do processo de segmentação os mapas de cores falsas obtidos foram comparados com a Histopatologia padrão a fim de avaliar a qualidade e consistência das imagens. Os resultados obtidos pela técnica de KMCA foram ligeiramente superiores com relação a HCA na identificação de pixels dos mapas morfo-quimicos correspondentes às estruturas histológicas da amostra. No entanto, nossos achados indicam que a escolha do substrato, dados de entrada, métodos de pré-processamento e preparação de amostras têm mais influência nos resultados finais do que o algoritmo de agrupamento usado para reconstruir as imagens. Na segunda fase do nosso estudo, a Análise de Componentes Principais (PCA) foi empregada para comparar os dados do grupo saudável aos animais expostos aos produtos carcinogênicos por 8, 16 e 48 semanas a fim de avaliar as alterações bioquímicas induzidas pela carcinogênese química. O desempenho da classificação em cada comparação pareada foi calculado usando um teste de classificação binária baseado na Análise de Discriminante Linear associada à técnica de PCA (PC-LDA). O método obteve discriminação satisfatória (acima de 80%) comparando tecido saudável com as amostras que foram classificadas como papiloma (16 semanas) e carcinoma espinocelular invasivo (48 semanas) independentemente do substrato usado para colocar as amostras. A comparação de pele saudável com animais expostos aos fatores carcinogênicos por 8 semanas (livres de malignidade de acordo com as evidências clínicas e morfológicas) apresentou figuras de performance cujos valores variaram entre 35-78%, indicando que a habilidade da técnica de PC-LDA em classificar corretamente dados espectrais de lesões cancerosas e pré-cancerosas variam com o estádio da doença durante o processo de tumorigênese. Diante disso, como uma prova de conceito, demonstramos a viabilidade da espectroscopia FTIR na avaliação dos eventos biológicos desencadeados pelo câncer usando uma metodologia que não requer colorações e substratos caros, assim como não interrompe/altera o fluxo de trabalho atual do patologista. Este é um passo importante na implementação da tecnologia no ambiente clínico, uma vez que o método pode ser usado para complementar o processo de diagnóstico do câncer como uma alternativa não-subjetiva e que não requer procedimentos trabalhosos e demorados, nem sondas caras como biomarcadores.

La présente thèse traite de la description de systèmes complexes, notamment des polymères et des cuprates, par la théorie de la fonctionnelle de la densité. En premier lieu, la théorie de la fonctionnelle de la densité ainsi que différentes fonctionnelles utilisées pour simuler les matériaux à l’étude sont présentées. Plus spécifiquement, les fonctionnelles LDA et GGA sont décrites et leurs limites sont exposées. De plus, le modèle de Hubbard ainsi que la fonctionnelle LDA+U qui en découle sont abordés dans ce chapitre afin de permettre la simulation des propriétés de matériaux à forte corrélation électronique. Par la suite, les résultats obtenus sur les polymères sont résumés par deux articles. Le premier traite de la variation de la bande interdite entre les polymères pontés et leurs homologues non pontés. Le second se penche sur l’étude de polymères à faible largeur de bande interdite. Dans ce dernier, il sera démontré qu’une fonctionnelle hybride, contenant de l’échange exact, est nécessaire afin de décrire les propriétés électroniques des systèmes à l’étude. Finalement, le dernier chapitre est consacré à l’étude des cuprates supraconducteurs. La LDA+U pouvant rendre compte de la forte localisation dans les orbitales 3d des atomes de cuivre, une étude de l’impact de cette fonctionnelle sur les propriétés électroniques est effectuée. Un dernier article investiguant différents ordres magnétiques dans le La2CuO4 dopé termine le dernier chapitre. On trouve aussi, en annexe, un complément d’information pour le second article et une description de la théorie de la supraconductivité de Bardeen, Cooper et Schrieffer. / Description of complex systems by Density functional theory is treated in this thesis. First, the Density functional theory and a few functionals used to simulate cristals are presented. Specifically, the LDA and GGA functionnals are described and their limits are exposed. Furthermore, the Hubbard model as well as the LDA+U functionnal are addressed in this chapter. These methods enable the study of highly correlated materials. Then, results obtained on polymers are summarized in two articles. The first one treats the band gap variation of ladder-type polymers compared to non ladder type ones. The second article considers small band gap polymers. In this case, it will be shown that an hybrid functional, which contains exact exchange, is required to describe the electronic properties of the polymers under study. Finally, the last chapter address the study of cuprates superconductors. The LDA+U can account for the localization of electrons in copper orbitals. Consequently, a study of the impact of this functionnal on electronic properties of cuprates is conducted. The chapter is ended by an article treating magnetic orders in doped La2CuO4. Supplementary materials of the second article and a description of the theory of superconductivity of Bardeen, Cooper and Schrieffer are put in annex.

La prise en compte du cadrage fait par les influenceurs de la droite radicale et du contenu de leur discours reste peu explorée. Ces contenus sont particulièrement préoccupants lorsqu’ils sont produits par des « influenceurs » qui auraient non seulement un pouvoir social sur leurs nombreux adeptes engagés, mais qui susciteraient aussi une opposition souvent virulente envers les autorités. Certains affirment que la recherche a manqué d’études empiriques systématiques sur le sujet et l’étude de la variation de cadre serait une piste intéressante pour de futures recherches (Benford 1997). Il y a donc un besoin pressant de développer une compréhension rigoureuse de la façon dont des crises mondiales peuvent changer la façon dont certains influenceurs de la droite radicale cadrent leurs discours. En utilisant des données originales sur cinq influenceurs de la droite radicale au Québec sur la plateforme Twitter de janvier 2020 à avril 2022, nous relevons d’abord les sujets prédominants dans le discours des influenceurs de la droite radicale. Grâce à une analyse thématique par LDA, nous confirmons que sept sujets dominent le discours des influenceurs de la droite radicale durant la pandémie de COVID-19, soit les élites, la gestion de crise, les médias, la fausse pandémie, la conspiration, le gouvernement et la liberté. Deuxièmement, nous montrons que la crise sanitaire de COVID-19 a poussé les influenceurs de la droite radicale à changer leur discours et à adopter trois « cadres de crise » qui présentent la COVID-19 comme directement liée aux concepts de gouvernance, de conspiration et de liberté. / The framing done by radical right influencers and the content of their discourse remain underexplored. Such content is of serious concern when it is produced by "influencers" who would not only have social power over their many committed followers, but also would generate often virulent opposition to the authorities. Some argue that research has lacked systematic empirical studies on the topic and the study of frame variation would be an interesting avenue for future research (Benford 1997). There is thus a pressing need to develop a rigorous understanding of how global crises can change the way some radical right-wing influencers frame their discourse. Using original data about five radical right influencers in Quebec on the Twitter platform from January 2020 to April 2022, we first identify the predominant topics in radical right influencers' discourse. Through a thematic analysis by LDA, we confirm that six topics dominate the discourse of radical right influencers during the pandemic of COVID-19: elites, crisis management, media, fake pandemic, conspiracy, and freedom. Second, we show that the COVID-19 health crisis pushed radical right influencers to change their discourse and adopt three "crisis frames" that present COVID-19 as directly related to the concepts of conspiracy, governance, and freedom.

Master of Science / Department of Computing and Information Sciences / Doina Caragea / William H. Hsu / In this work, we focus on the task of clustering businesses in the state of Kansas based on the content of their websites and their business listing information. Our goal is to cluster the businesses and overcome the challenges facing current approaches such as: data noise, low number of clustered businesses, and lack of evaluation approach. We propose an LSA-based approach to analyze the businesses’ data and cluster those businesses by using Bisecting K-Means algorithm. In this approach, we analyze the businesses’ data by using LSA and produce businesses’ representations in a reduced space. We then use the businesses’ representations to cluster the businesses by applying the Bisecting K-Means algorithm. We also apply an existing LDA-based approach to cluster the businesses and compare the results with our proposed LSA-based approach at the end. In this work, we evaluate the results by using a human-expert-based evaluation procedure. At the end, we visualize the clusters produced in this work by using Google Earth and Tableau. According to our evaluation procedure, the LDA-based approach performed slightly bet- ter then the LSA-based approach. However, with the LDA-based approach, there were some limitations which are: low number of clustered businesses, and not being able to produce a hierarchical tree for the clusters. With the LSA-based approach, we were able to cluster all the businesses and produce a hierarchical tree for the clusters.

O uso de dados biométricos da face para verificação automática de identidade é um dos maiores desafios em sistemas de controle de acesso seguro. O processo é extremamente complexo e influenciado por muitos fatores relacionados à forma, posição, iluminação, rotação, translação, disfarce e oclusão de características faciais. Hoje existem muitas técnicas para se reconhecer uma face. Esse trabalho apresenta uma investigação buscando identificar uma face no banco de dados ORL com diferentes grupos de treinamento. É proposto um algoritmo para o reconhecimento de faces baseado na técnica de subespaço LDA (PCA + LDA) utilizando uma rede neural SOM para representar cada classe (face) na etapa de classificação/identificação. Aplicando o método do subespaço LDA busca-se extrair as características mais importantes na identificação das faces previamente conhecidas e presentes no banco de dados, criando um espaço dimensional menor e discriminante com relação ao espaço original. As redes SOM são responsáveis pela memorização das características de cada classe. O algoritmo oferece maior desempenho (taxas de reconhecimento entre 97% e 98%) com relação às adversidades e fontes de erros que prejudicam os métodos de reconhecimento de faces tradicionais. / The use of biometric technique for automatic personal identification is one of the biggest challenges in the security field. The process is complex because it is influenced by many factors related to the form, position, illumination, rotation, translation, disguise and occlusion of face characteristics. Now a days, there are many face recognition techniques. This work presents a methodology for searching a face in the ORL database with some different training sets. The algorithm for face recognition was based on sub-space LDA (PCA + LDA) technique using a SOM neural net to represent each class (face) in the stage of classification/identification. By applying the sub-space LDA method, we extract the most important characteristics in the identification of previously known faces that belong to the database, creating a reduced and more discriminated dimensional space than the original space. The SOM nets are responsible for the memorization of each class characteristic. The algorithm offers great performance (recognition rates between 97% and 98%) considering the adversities and sources of errors inherent to the traditional methods of face recognition.

Live cell imaging with FTIR spectroscopy offers a high throughput, non-damage and lab-free method to study the cells in vivo which has significant advantages in the field of cancer diagnosis and drug screening. However, due to the strong absorbance of water, using infrared spectroscopy in such field remains to be an underdeveloped topic. This project demonstrates a novel method to perform IR imaging of cells in solution. A novel water correction method, which avoids the using of water combination band, is proposed. A buffer reference and a cell reference spectra were introduced to fitting the contribution based on protein bands. This method was implemented on three types of IR spectrometers, namely conventional FTIR spectrometer, synchrotron-based FTIR spectrometer and quantum cascade laser (QCL) microscope. To date, most of the live cell imaging carried out with IR sources utilise synchrotron radiation. Recently, a new bench top system, QCL microscope, has been developed. It incorporates four tunable QCL laser sources covering the wavenumber range 900-1800 cm-1 which are many orders of magnitude brighter than conventional sources. The proposed water correction method is, therefore, capable of processing the data recorded by all three types of IR spectrometers. Three prostate cancer cell lines were employed to evaluate the water correction method and the performance of three spectrometers on imaging of cell in solution. The obtained spectra was analysed with multivariate analysis, PCA and PC-LDA which shows good separation between cell lines. The data was also examined with Random Forest algorithm to establish a classifier and the diagnostic capability of the water corrected spectra was proven.

Prior research on text collections of religious documents has demonstrated that viable recommender systems in the area are lacking, if not non-existent, for some datasets. For example, both www.LDS.org and scriptures.byu.edu are websites designed for religious use. Although they provide users with the ability to search for documents based on keywords, they do not provide the ability to discover documents based on similarity. Consequently, these systems would greatly benefit from a recommender system. This work provides a framework for evaluating recommender systems and is flexible enough for use with either website. Such a framework would identify the best recommender system that provides users another way to explore and discover documents related to their current interests, given a starting document. The framework created for this thesis, RelRec, is attractive because it compares two different recommender systems. Documents are considered relevant if they are among the nearest neighbors, where "nearest" is defined by a particular system's similarity formula. We use RelRec to compare output of two particular recommender systems on our selected data collection. RelRec shows that LDA recommeder outperforms the TF-IDF recommender in terms of coverage, making it preferable for LDS-based document collections.

It has been noticed for a long time that humans can identify others based on their biological movement from a distance. However, it is only recently that computer vision based gait biometrics has received much attention. In this dissertation, we perform a thorough study of gait recognition from a computer vision perspective. We first present a parameterless baseline recognition algorithm, which bases similarity on spatio-temporal correlation that emphasizes gait dynamics as well as gait shapes. Our experiments are performed with three popular gait databases: the USF/NIST HumanID Gait Challenge outdoor database with 122 subjects, the UMD outdoor database with 55 subjects, and the CMU Mobo indoor database with 25 subjects. Despite its simplicity, the baseline algorithm shows strong recognition power. On the other hand, the outcome suggests that changes in surface and time have strong impact on recognition with significant drop in performance. To gain insight into the effects of image segmentation on recognition -- a possible cause for performance degradation, we propose a silhouette reconstruction method based on a Population Hidden Markov Model (pHMM), which models gait over one cycle, coupled with an Eigen-stance model utilizing the Principle Component Analysis (PCA) of the silhouette shapes. Both models are built from a set of manually created silhouettes of 71 subjects. Given a sequence of machine segmented silhouettes, each frame is matched into a stance by pHMM using the Viterbi algorithm, and then is projected into and reconstructed by the Eigen-stance model. We demonstrate that the system dramatically improves the silhouette quality. Nonetheless, it does little help for recognition, indicating that segmentation is not the key factor of the covariate impacts. To improve performance, we look into other aspects. Toward this end, we propose three recognition algorithms: (i) an averaged silhouette based algorithm that deemphasizes gait dynamics, which substantially reduces computation time but achieves similar recognition power with the baseline algorithm; (ii) an algorithm that normalizes gait dynamics using pHMM and then uses Euclidean distance between corresponding selected stances -- this improves recognition over surface and time; and (iii) an algorithm that also performs gait dynamics normalization using pHMM, but instead of Euclidean distances, we consider distances in shape space based on the Linear Discriminant Analysis (LDA) and consider measures that are invariant to morphological deformation of silhouettes. This algorithm statistically improves the recognition over all covariates. Compared with the best reported algorithm to date, it improves the top-rank identification rate (gallery size: 122 subjects) for comparison across hard covariates: briefcase, surface type and time, by 22%, 14%, and 12% respectively. In addition to better gait algorithms, we also study multi-biometrics combination to improve outdoor biometric performance, specifically, fusing with face data. We choose outdoor face recognition, a "known" hard problem in face biometrics, and test four combination schemes: score sum, Bayesian rule, confidence score sum, and rank sum. We find that the recognition power after combination is significantly stronger although individual biometrics are weak, suggesting another effective approach to improve biometric recognition. The fundamental contributions of this work include (i) establishing the "hard" problems for gait recognition involving comparison across time, surface, and briefcase carrying conditions, (ii) revealing that their impacts cannot be explained by silhouette segmentation, (iii) demonstrating that gait shape is more important than gait dynamics in recognition, and (iv) proposing a novel gait algorithm that outperforms other gait algorithms to date.

Dans cette thèse on s'intéresse aux propriétés spectrales et optiques des matériaux aux corrélations fortes de Coulomb. En introduisant un schéma pour la continuation analytique des énergies propres de Matsubara sur l'axe réel dans le cadre le plus général de la théorie de champ moyen dynamique des clusters, les propriétés spectrales résolues en angles deviennent accessibles. De plus, dans certaines approximations, ceci permet la détermination des propriétés optiques. Notamment, on développe dans cette thèse un formalisme pour la conductivité optique dans le langage d'une base localisée. En généralisant la substitution de Peierls pour les cas réalistes de plusieurs atomes par cellule élémentaire, nous obtenons une approche très flexible et facile a implémenter. Ces techniques sont appliquées à plusieurs composés: *l'oxyde de vanadium VO2* Alors que sa phase métallique est caractérisée par des forts effets de corrélations dynamiques, le spectre du VO2 isolant est bien décrit dans un langage à un particule, que l'on définira. L'image de l'isolant qui se forme à partir de notre analyse est celle d'un "isolant de Peierls à N particules". De plus, on calcule la conductivité optique des deux phases et on trouve un accord satisfaisant avec des expériences récentes. Finalement, on en déduit aussi la couleur du matériau. *les sesqui-oxyde des terres rares RE2O3 (RE=Ce, Pr, Nd, Pm)* Dans cette série on trace l'influence des orbitales 4f dans la bande interdite d'excitation et dans les spectres optiques. On trouve un accord raisonnable avec des expériences. *le sesqui-oxyde de vanadium V2O3" Dans notre analyse on trouve que la transition métal-isolant en fonction du dopage avec du chromium est causée par un agrandissement de l'écart due au champ cristallin, résultante des corrélations. Egalement, on trouve une sélectivité orbitale de la cohérence des quasi-particules dans le métal, qui permet à expliquer des résultats expérimentaux optiques récentes.