Cette recherche vise à étudier l'efficacité des cours d'espagnol comme langue étrangère, à partir de la perception des étudiants d’espagnol comme langue d’origine. Pour atteindre les résultats et les conclusions obtenus, cette étude qualitative compte sur la participation des répondants hispaniques qui ont appris l'espagnol à la maison, à l'école primaire, au secondaire, au cégep et/ou au collège. Les participants ont répondu à un questionnaire qui leur a permis d'exprimer leur perception à l'égard des cours d’espagnol comme langue étrangère qu'ils ont pris au cours de leur étude primaire, secondaire ou supérieure. Les résultats de cette recherche indiquent que les élèves hispaniques ne sont pas intégrés dans les classes ELE, ce qui motive les participants à poursuivre l’étude de la langue espagnole. Comme expliqué par Gardner (1985) dans sa théorie du modèle socio-éducationnel, l’acquisition des langages intervient dans un contexte culturel spécifique et elle est influencée par la croyance populaire et les différences individuelles. Ces dernières sont intégrées dans le concept du motif d'intégration, qui est composé du sens de l'intégration, de l'attitude envers la situation d'apprentissage et de la motivation. Les réponses données par les participants dans cette étude peuvent être utiles à la conception et au développement de cours et de méthodes d’apprentissage de l’espagnol pour les hispanophones d’origine. En tenant compte de leurs points de vue, les étudiantes auraient un sens plus élevé de l’intégration, ce qui influence directement leur motivation et par le fait même leur processus d’apprentissage. / The purpose of this research is to study the efficiency of Spanish courses from a Heritage Learner’s perspective. This qualitative study examines Hispanics who have learned Spanish at home, in elementary school, in high school, in CEGEP and/or in university. The participants answered a questionnaire that allowed them to express their perception of the Spanish courses that they had taken in their past. The results of this research indicate that L1/L2 users are not integrated in the Spanish classes. Participants demonstrate a reduced sense of motivation for the continuation of these types of courses. Gardner (1985) explains in his socio-educational model theory that language acquisition occurs in a specific cultural context and is influenced by cultural beliefs and individual differences. The latter are integrated in the concept of integrative motive, which is composed of the integrativeness, the attitude toward the learning situation and motivation. The answers given by the participants of this study may be of great help in the design and development of Spanish courses and books for L1/L2 users. By considering their views, students would have a higher sense of integration; an important factor that motivates students to pursue the study of Spanish. / Esta investigación tiene como objetivo final estudiar la eficacia de los cursos de ELE desde la percepción de los usuarios de L1/L2, o aprendices de herencia. Para llegar a los resultados y las conclusiones obtenidos, se realizó un estudio cualitativo que contó con la participación de informantes de origen hispano que han aprendido el español en casa, en la escuela primaria, en la secundaria, en el Cégep y/o en la universidad. Los participantes contestaron un cuestionario que les permitió expresar su percepción con respecto a los cursos de ELE, cursos que habían tomado durante sus estudios primarios, secundarios o superiores. Los resultados de esta investigación demuestran que los estudiantes de L1/L2 no se encuentran integrados en las clases de ELE, hecho que desmotiva a los informantes para continuar en el estudio de la lengua española. Tal y como explica Gardner (1985) en su teoría del Modelo socioeducacional, la adquisición de lenguas ocurre en un contexto cultural determinado, y está influenciada por la creencia popular y por las diferencias individuales. Estas se encuentran integradas en el concepto de motivo integrador, que está compuesto por el sentido de integración, la actitud hacia la situación de aprendizaje y la motivación. Las respuestas dadas por los informantes pueden servir de ayuda para el diseño y la creación de cursos y métodos de español destinados a usuarios de L1/L2. Al tomar en cuenta sus opiniones, los estudiantes tendrían un sentido de integración más alto, hecho que influye directamente en la motivación de los mismos y, por tanto, en su proceso de aprendizaje.

Tikhonov-type regularization of linear and nonlinear ill-posed problems in abstract spaces under sparsity constraints gained relevant attention in the past years. Since under some weak assumptions all regularized solutions are sparse if the l1-norm is used as penalty term, the l1-regularization was studied by numerous authors although the non-reflexivity of the Banach space l1 and the fact that such penalty functional is not strictly convex lead to serious difficulties. We consider the case that the sparsity assumption is narrowly missed. This means that the solutions may have an infinite number of nonzero but fast decaying components. For that case we formulate and prove convergence rates results for the l1-regularization of nonlinear operator equations. In this context, we outline the situations of Hölder rates and of an exponential decay of the solution components.

Various studies have indicated during the past decade that language transfer in L3 may not only stem from L1 but from L2 as well, and that it might sometimes even be stronger from L2, depending on certain factors that facilitate or inhibit transfer. This phenomenon of L2 as the main transfer source in L3 has often been referred to as the ”L2 status factor” (Hammarberg, 2001). The L2 status factor hypothesis expects that the priorly acquired language which scores the highest in several transfer factors will adopt the role of ”external supplier language” (Hammarberg, 2001), i.e. it will be the main source of transfer providing L3 with linguistic features. Namely, the factors that have hitherto been proposed to condition transfer are: typology, psychotypology, proficiency, and psychoaffective factors. The aim of this investigation is to compare the transfer that two groups might exhibit with regard to the English existential expletive pronoun (there), in order to account to whether transfer in L3 might be stronger from L2 than from L1 in this syntactic context. One group consists of subjects with different L1s, L2 Swedish and L3 English; the other is formed by L1 Swedish and L2 English speakers. The informants are aged 13-14, speak the L1s and Swedish (nearly) fluently and English at a basic/intermediate level. Basing the study on the L2 status factor hypothesis, and taking several transfer factors into consideration when analyzing the collected written data, the results are discussed both from a general perspective (from aggregate group scores) and from a micro-perspective (by tracing individual differences). The results obtained suggest that transfer in L3 appears to be stronger from L2 than from L1 when evaluating the aggregate group scores, but only in some cases (and not in most) when examining the individuals separately.

La liaison en français consiste en l’apparition d’une consonne de liaison entre deux mots dans des contextes précis de la chaîne parlée qui sont divisés en trois catégories : les contextes où la liaison est catégorique (un /n/ enfant), ceux où elle est variable (un gros /z/ avion) et ceux où elle est erratique (une chanson // anglaise). L’objectif de cette recherche est double : dans la première partie, nous exposons une analyse de chaque dimension linguistique et sociolinguistique impliquée dans le phénomène de la liaison. Chacune d’entre elles permet d’éclairer les études qui ont été réalisées ces dernières années auprès d’apprenants de FLE de niveau intermédiaire à avancé, afin de découvrir dans quelle mesure ils parviennent à s’approprier la liaison, en perception comme en production. Dans la seconde partie, nous présentons une étude expérimentale de l’appropriation de la liaison en français suivie par 17 apprenants coréens de FLE de niveau A1-A2 du CECRL enregistrés trois fois sur une durée d’un an, ainsi que par 165 enfants francophones natifs de deux à six ans. Comme il est impossible de comparer les deux groupes de participants du fait de conditions environnementales et de compétences cognitives non alignables, nous avons élaboré une méthode innovante d’appariement sur la base de données factuelles. Nos résultats nous permettent non seulement d’étayer le modèle constructionniste de l’acquisition de la liaison L1 (Chevrot, Dugua, & Fayol, 2009; Chevrot, Dugua, Harnois-Delpiano, Siccardi, & Spinelli, 2013) mais aussi d’esquisser le processus d’appropriation de la liaison L2, très influencé par la graphie apprise dès les premiers cours de FLE. Nous concluons par une mise en perspective didactique prenant en compte l’ensemble des prismes de ce kaléidoscope linguistique de la liaison en français car sans une perception claire de chacun d’entre eux, les apprenants de FLE ne peuvent qu’avoir une vision floue du phénomène. / French liaison is the appearance of a liaison consonant between two words in specific contexts within the spoken word chain. These contexts are divided into three categories : those where the liaison is categorical (e.g. “a child” : un /n/ enfant ), those where it is variable (e.g. “a big plane” : un gros /z/ avion) and those where it is forbidden (e.g. “an english song” : une chanson // anglaise ).The aim of this research is twofold. In the first part, we present an analysis of each linguistic and sociolinguistic sub-field involved in the phenomenon of French liaison. Each of them will allow to shed light on studies of the acquisition of French liaison that have been conducted with learners of FFL (intermediate to advanced level) in recent years. In the second part, we present a comparative study of the acquisition of French liaison based on experimental tasks performed by 17 Korean learners of FFL (level A1-A2 of the CEFR) recorded three times over a one-year and by 165 native children aged two to six. As it is impossible to compare the two groups of participants due to environmental conditions and to different cognitive skills, we have developed an innovative matching method based on factual data. On the one hand, this supports the constructionist model of the acquisition of the liaison L1 (Chevrot, Dugua, & Fayol, 2009; Chevrot, Dugua, Harnois-Delpiano, Siccardi, & Spinelli, 2013). On the other hand, this enables us to sketch the acquisition process of the liaison L2, very influenced by the spelling which is learned from the first FFL course. We conclude with a didactic perspective layout taking into account all the prisms of this linguistic kaleidoscope of French liaison because without a clear perception of each one, FFL learners are bound to have a partial and unfocused interpretation of this phenomenon.

The Human Papillomavirus (HPV) is a virus of the Papillomaviridae family and its genome consists of DNA. It is the main cause of viral sexual transmission. Cervical cancer or uterine colon cancer is the second most common occurrence among women worldwide. Recent studies have proven that some types of HPV are mainly responsible for the development of this type of cancer. In accordance to its oncogenic activity this viral group was divided in low and high risk types. Brazil does not have a representative amount of data related to the prevalence of HPV infection. The incidence data of the virus is obtained through the analysis of carriers of the invasive carcinoma of uterine colon, cervical intraepithelial neoplasias, and others types of infections associated. Molecular assays have been showing throughout the years an excellent sensitivity and specificity in the detection of the HPV s DNA. Nowadays, in situ hybridization techniques are being used as method of choice for the detection of the viral DNA. In many studies, the Polymerase Chain Reaction (PCR) with the application of the primers MY09 and MY11 have shown efficiency in viral tracking, consequently, the development of molecular diagnostics allow monitoring of the disease, decreasing mortality rates caused by cancer of the cervix. The DNA sequencing is an efficient methodology used for the molecular characterization of the viral types, which allows the accomplishment of the alignment for similarity of the sequences obtained through others that were already identified and deposited in the GenBank. The current study s purpose was to identify the different types of HPV and the presence of co-infections through PCR and sequencing of a hyper-variable region of the L1 gene. The studied sample consisted of 515 female volunteers, which 111 (21.55%) presented the incidence of the HPV DNA. Within the acquired specimens, 50% were considered high oncogenic risk (the major incidence was types 16 and 58) and 7.21% had multiple infection. The determination of the base sequencing of the L1 gene is essential to the identification of the viral type. However, the sequencing of the complete gene and both DNA chains become financially unviable for the majority of the specialized laboratories. A practical approach applied in this study was to determine the sequencing of a specific region of the L1 gene with 450pb, where the identification of the HPV became possible by following this methodology. The techniques used showed the necessity of the application of a more sensible and specific viral diagnosis, which contributes to the viral infection control and to the reduction of the mortality rate caused by cervical cancer. / Fundação de Amparo a Pesquisa do Estado de Alagoas / O Papilomavírus Humano (HPV) é um vírus da família Papillomaviridae e possui o seu genoma constituído de DNA. É o causador da transmissão sexual viral de maior frequência. O câncer cervical ou câncer de colo de útero possui a segunda maior ocorrência nas mulheres de todo o mundo. Estudos recentes têm comprovado que alguns tipos de HPV são os principais responsáveis pelo desenvolvimento deste tipo de câncer. De acordo com a sua atividade na carninogênese esse grupo viral foi dividido em tipos de baixo e de alto risco oncogênico. O Brasil ainda não possui uma quantidade representativa de dados relacionados à prevalência de infecção por HPV, os dados de incidência do vírus são obtidos através da análise de portadores de carcinoma invasivo de colo uterino, neoplasias intraepiteliais cervicais e outros tipos de infecções associadas. Ensaios moleculares vêm mostrando ao longo dos anos uma alta sensibilidade e especificidade na detecção do DNA do HPV. Atualmente são utilizadas técnicas de hibridização como método de escolha para a detecção do DNA viral. Em muitos estudos a Reação em Cadeia da Polimerase (PCR) com a aplicação dos primers MY09 e MY11 demonstrou eficiência no rastreamento viral, consequentemente, o desenvolvimento desse diagnóstico molecular possibilitaria o monitoramento da doença, diminuindo as taxas de mortalidade causada pelo câncer de colo do útero. O sequenciamento de DNA é uma metodologia eficiente para a caracterização molecular dos tipos virais, permitindo a realização do alinhamento por similaridade das sequencias obtidas com outras já identificadas e depositadas no GenBank. O presente trabalho teve como propósito identificar os diferentes tipos de HPV e a presença de co-infecções, através de PCR e sequenciamento de uma região hipervariável do gene L1. A amostra estudada foi composta de 515 voluntárias das quais 111 (21,55%) apresentaram a presença do DNA do HPV. Entre os espécimes encontrados 50% são considerados de alto risco oncogênico (maior incidência dos tipos 16 e 58) e 7,21% possuíam infecção múltipla. A determinação das sequencias de base do gene L1 é fundamental para a identificação do tipo viral. Entretanto o sequenciamento do gene completo e em ambos os sentidos da cadeia de DNA se tornam inviáveis financeiramente para a maioria dos laboratórios especializados. Uma abordagem prática aplicada no estudo foi determinar a sequencia de uma região específica do gene L1 contendo 450pb, sendo possível a identificação do HPV seguindo essa metodologia. As técnicas utilizadas mostraram a necessidade da aplicação de um diagnóstico viral mais sensível e específico, contribuindo para o controle da infecção viral e para a diminuição da incidência e da mortalidade causadas pelo câncer cervical.

Submitted by Erika Demachki (erikademachki@gmail.com) on 2015-01-16T19:06:55Z No. of bitstreams: 2 Tese - Adriana da Silva Santos - 2014.pdf: 3299451 bytes, checksum: 52a612ede01e4713d7dd458d61176fe1 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Erika Demachki (erikademachki@gmail.com) on 2015-01-16T19:07:03Z (GMT) No. of bitstreams: 2 Tese - Adriana da Silva Santos - 2014.pdf: 3299451 bytes, checksum: 52a612ede01e4713d7dd458d61176fe1 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-01-16T19:07:03Z (GMT). No. of bitstreams: 2 Tese - Adriana da Silva Santos - 2014.pdf: 3299451 bytes, checksum: 52a612ede01e4713d7dd458d61176fe1 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2014-05-30 / Conselho Nacional de Pesquisa e Desenvolvimento Científico e Tecnológico - CNPq / Swine leptospirosis is a worldwide anthropozoonosis caused by spirochetes of the species Leptospira interrogans, predominantly by serovars Pomona, Icterohaemorrhagiae, Taravossi, Canicola Grippotyphosa and Bratislava. Experimental infection were performed in 18 sows at 40 days of age with L. interrogans serogroup Icterohaemorrhagiae serovar Conpenhageni strain Fiocruz L1-130, and serogroup Pomona serovar Kennewicki strain Pomona From divided in groups: GI (control), GII (inoculated with serogroup Pomona) and GIII (inoculated with serogroup Icterohaemorrhagiae). Clinical course of the disease, results of serological tests and detection of bacteria by polymerase chain reaction (PCR) and bacterial isolation during the experiment, were compared between groups. Additionally, the main histological changes observed in brain, liver, lymph nodes, lung and kidney of 12 animals experimentally infected and seropositive for leptospirosis were described. In lesions suggestive of the disease the technique of silver impregnation of Warthin-Starry, was used as direct detection method. During the experimental period, all samples of the GI (control) were negative in the microscopic agglutination test, with the GII positivity exclusive to serogroup Pomona in six animals from 6 dpi, and in GIII, only to serogroup Icterohaemorrhagiae in five animals from 3 dpi. At necropsy the only change was pale kidney in 50% of GI, 83.33% of GII and 83.33% of GIII (p> 0.05). Only in seropositive pigs was found lesions suggestive of leptospirosis as focal microgliosis in the brain (8.33%), lymphoid hyperplasia in lymph nodes of the mediastinum a nd mesentery (41.66%), mononuclear interstitial pneumonia (58.33 %), multifocal mononuclear hepatitis, further acute tubular nephrosis (100%) and interstitial nephritis (91.66%). There was a strong association between serological condition and lesions of mononuclear hepatitis (p = 0.0128), mononuclear interstitial nephritis (p = 0.0004), tubular nephrosis (p <0.0001) and mononuclear interstitial pneumonia (p = 0, 0377). / A leptospirose suína é uma antropozoonose difundida mundialmente e causada por espiroquetas da espécie Leptospira interrogans, principalmente pelos sorovares Pomona, Icterohaemorrhagiae, Taravossi, Canicola, Grippotyphosa e Bratislava. Neste trabalho descreve-se a infecção experimental por L. interrogans sorogrupo Icterohaemorrhagiae sorovar Conpenhageni estirpe Fiocruz L1 -130, e sorogrupo Pomona sorovar Kennewicki estirpe Pomona From, em 18 fêmeas suínas com 40 dias de idade divididas em três grupos: GI (controle), GII (inoculados com sorogrupo Pomona) e GIII (inoculados com sorogrupo Icterohaemorrhagiae). Compararam -se entre os grupos, o curso clínico da doença, os resultados dos exames sorológicos e de detecção da bactéria durante o período experimental pelas técnicas de reação em cadeia da polimerase (PCR) e isolamento bacteriano. Foram ainda descritas as principais alterações histológicas observadas em cérebro, fígado, linfonodos, pulmão e rins de 12 animais experimentalmente infectados e soropositivos para leptospirose. Em lesões sugestivas da doença utilizou-se a técnica de impregnação pela prata de Warthin-Starry, como método de detecção direta. Durante o período experimental foram negativas na soroaglutinação microscópica todas as amostras do GI (controle), havendo no GII positividade exclusiva ao sorogrupo Pomona nos seis animais a partir do 6º dia pós-infecção (dpi), e no GIII, apenas ao sorogrupo Icterohaemorrhagiae, em cinco animais a partir do 3º dpi. Na necropsia a única alteração observada foi rim pálido em 50% do G I, 83,33% do GII e em 83,33% GIII (p>0,05). Somente nos suínos soropositivos constataram-se lesões sugestivas de leptospirose, como microgliose mononuclear focal no cérebro (8,33%), hiperplasia linfoide em linfonodos do mediastino e mesentério (41,66%), pneumonia intersticial mononuclear multifocal (58,33%), hepatite mononuclear focal (41,66%), além de nefrose tubular aguda (100%) e nefrite intersticial (91,66%). Houve forte associação entre a condição sorológica e as lesões de hepatite mononuclear (p=0,0128), nefrite intersticial mononuclear (p=0,0004), nefrose tubular (p<0,0001) e pneumonia intersticial mononuclear (p=0,0377).

Cette thèse renseigne sur la structuration du discours d’écoliers haïtiens à des stades spécifiques de leur acquisition du français langue de scolarisation. A partir de tâche narrative de construction de récit, elle étudie d’une part leur capacité à mettre en mots des événements complexes, à produire un récit structuré et cohérent (analyse macro-structurelle) et, d’autres part, les moyens référentiels qu’ils mobilisent pour introduire, maintenir, réintroduire les protagonistes l’histoire à raconter (analyse micro-structurelle). Les récits ont été recueillis en créole haïtien L1 et en français L2 à partir de la planche narrative les Oisillons. Le corpus est constitué de 160 récits (80 en créole haïtien, 80 en français). Les enquêtés, d’âge et de niveau scolaire différents (9-10 ans/4ème année et de 11-12 ans/6ème année) viennent de 4 écoles différentes de la capitale haïtienne, positionnées différemment sur l’échelle des valeurs sociale et scolaire. L’âge, le niveau scolaire, le contexte d’appropriation du français ont été considérés pour les deux axes d’analyse. Les investigations portent essentiellement sur les récits produits en français L2, langue en cours d’acquisition mais certains éléments sont examinés au regard de la L1. Les résultats d’analyse révèlent principalement une variation importante dans le développement de la capacité narrative et linguistique des sujets en L2 entre les quatre groupes scolaires représentés. Ils montrent par la même occasion l’influence de l’école, lieu principal d’acquisition du français sur le développement des compétences linguistiques des écoliers haïtiens en L2, aspect qui est décrit dans cette présente étude. / This doctoral dissertation provides information on how Haitian pupils structure their written text at specific stages of the process of acquiring French as their academic language. Examining how narrative stories are constructed, on the one hand, the capacity of students to express complex events (macro-structural analysis) and, on the other hand, the referential means that are put to work: introducing, maintaining, and reintroducing the protagonists and the story to tell (micro-structural analysis). Stories have been gathered in Haitian Creole L1 and in French L2 from the story les Oisillons (Young birds). The corpus is made up of 160 stories (80 in Haitian Creole, 80 in French). The pupils surveyed whose age and school level are different (9-10 years old / 4th grade and 11-12 years old / 6th grade) come from 4 different schools of the Haitian capital. These schools occupy different positions on the scale of social and school values. Several factors including age, school level, and acquisition context of the French language have been considered for the two axes of analysis. The research focuses mainly on stories written in French L2, which is the language in the process of being acquired, but some elements of L1 are also examined. The results of the analysis reveal mainly an important variation in the development of the narrative and language capacity of the subjects in L2 between the 4 school groups that are represented. At the same time, it is shown how school which is the main place for the acquisition of French influences the development of language competence of Haitian pupils in L2. This is the point that is described in this study.

This thesis explores the role of age in second language (L2) acquisition and first language (L1) attrition. The focus is on Voice Onset Time (VOT) in the production and categorical perception of word-initial L1 and L2 stops in highly advanced L1 Spanish learners of L2 Swedish. Using as the point of departure a maturational constraints perspective and the Critical Period Hypothesis (CPH), Study I examines the impact of age of onset (AO) of L2 acquisition on the production of L2 Swedish voiceless stops. The results show that there are AO effects even in the speech of highly advanced L2 learners and that the incidence of nativelike L2 learners is considerably lower than earlier assumed. However, conclusions like these are only possible when speaking rate is accounted for, thereby highlighting the importance of speaking rate effects on VOT as a measure of nativelikeness. Like Study I, Study II reveals age effects on the same L2 learners’ categorical perceptions of L2 Swedish stops. Moreover, after combining the results with the data from Study I, the incidence of nativelike behavior drops remarkably with no late L2 learner performing within the range of native-speaker production and perception. The results suggest that L2 acquisition of phonetic/phonological aspects is especially sensitive to AO effects. It is concluded that theories on maturational constraints, including the CPH, cannot be refuted on the basis of the present data. Study III concerns the same participants’ production and perception of L1 Spanish stops. Age of reduced contact (ARC) is identified as an important predictor for L1 attrition and retention of voiceless stop production, although not of stop perception. This discrepancy is related to different activation thresholds as proposed by the Activation Threshold Hypothesis (ATH). It is further suggested that early bilinguals are more dependent on high-frequency L1 use than late bilinguals when compensating for age effects, but only in production. / <p>At the time of doctoral defence the following papers were unpublished and had a status as follows: Paper 1: Accepted. Paper 2: In press. Paper 3: Manuscript.</p> / Age of onset and ultimate attainment in second language acquisition, The Bank of Sweden Tercentenary Foundation, grant no. 1999-0383:01 / First language attrition in advanced second language speakers, Swedish Research Council, grant no. 421-2004-1975

Dans les maisons et bâtiments intelligents, il devient nécessaire de limiter l'intervention humaine sur le système énergétique, afin de fluctuer automatiquement l'énergie consommée par les appareils consommateurs. Pour cela, un système de mesure de la consommation électrique d'équipements est aussi nécessaire et peut être déployé de deux façons : intrusive ou non-intrusive. La première solution consiste à relever la consommation de chaque appareil, ce qui est inenvisageable à une grande échelle pour des raisons pratiques liées à l'entretien et aux coûts. Donc, la solution non-intrusive (NILM pour Non-Intrusive Load Monitoring), qui est capable d'identifier les différents appareils en se basant sur les signatures extraites d'une consommation globale, est plus prometteuse. Le problème le plus difficile des algorithmes NILM est comment discriminer les appareils qui ont la même caractéristique énergétique. Pour surmonter ce problème, dans cette thèse, nous proposons d'utiliser une information externe pour améliorer la performance des algorithmes existants. Les premières informations additionnelles proposées considèrent l'état précédent de chaque appareil comme la probabilité de transition d'état ou la distance de Hamming entre l'état courant et l'état précédent. Ces informations sont utilisées pour sélectionner l'ensemble le plus approprié des dispositifs actifs parmi toutes les combinaisons possibles. Nous résolvons ce problème de minimisation en norme l1 par un algorithme d'exploration exhaustive. Nous proposons également d'utiliser une autre information externe qui est la probabilité de fonctionnement de chaque appareil fournie par un réseau de capteurs sans fil (WSN pour Wireless Sensor Network) déployé dans le bâtiment. Ce système baptisé SmartSense, est différent de la solution intrusive car seul un sous-ensemble de tous les dispositifs est surveillé par les capteurs, ce qui rend le système moins intrusif. Trois approches sont appliquées dans le système SmartSense. La première approche applique une détection de changements de niveau sur le signal global de puissance consommé et les compare avec ceux existants pour identifier les dispositifs correspondants. La deuxième approche vise à résoudre le problème de minimisation en norme l1 avec les algorithmes heuristiques de composition Paréto-algébrique et de programmation dynamique. Les résultats de simulation montrent que la performance des algorithmes proposés augmente significativement avec la probabilité d'opération des dispositifs surveillés par le WSN. Comme il n'y a qu'un sous-ensemble de tous les appareils qui sont surveillés par les capteurs, ceux qui sont sélectionnés doivent satisfaire quelques critères tels qu'un taux d'utilisation élevé ou des confusions dans les signatures sélectionnées avec celles des autres. / In smart homes, human intervention in the energy system needs to be eliminated as much as possible and an energy management system is required to automatically fluctuate the power consumption of the electrical devices. To design such system, a load monitoring system is necessary to be deployed in two ways: intrusive or non-intrusive. The intrusive approach requires a high deployment cost and too much technical intervention in the power supply. Therefore, the Non-Intrusive Load Monitoring (NILM) approach, in which the operation of a device can be detected based on the features extracted from the aggregate power consumption, is more promising. The difficulty of any NILM algorithm is the ambiguity among the devices with the same power characteristics. To overcome this challenge, in this thesis, we propose to use an external information to improve the performance of the existing NILM algorithms. The first proposed additional features relate to the previous state of each device such as state transition probability or the Hamming distance between the current state and the previous state. They are used to select the most suitable set of operating devices among all possible combinations when solving the l1-norm minimization problem of NILM by a brute force algorithm. Besides, we also propose to use another external feature that is the operating probability of each device provided by an additional Wireless Sensor Network (WSN). Different from the intrusive load monitoring, in this so-called SmartSense system, only a subset of all devices is monitored by the sensors, which makes the system quite less intrusive. Two approaches are applied in the SmartSense system. The first approach applies an edge detector to detect the step-changes on the power signal and then compare with the existing library to identify the corresponding devices. Meanwhile, the second approach tries to solve the l1-norm minimization problem in NILM with a compositional Pareto-algebraic heuristic and dynamic programming algorithms. The simulation results show that the performance of the proposed algorithms is significantly improved with the operating probability of the monitored devices provided by the WSN. Because only part of the devices are monitored, the selected ones must satisfy some criteria including high using rate and more confusions on the selected patterns with the others.

The master's thesis deals with the restoration of audio signals that have been damaged by clipping. Used methods are based on sparse representations of signals. The introduction of the thesis explains the issue of clipping and mentions the list of already existing methods that solve declipping, which are followed by the thesis. In the next chapter, the necessary theory of sparse representations and the proximal algorithms is described, including specific representatives from the category of convex optimization problems. The thesis contains declipping algorithm implemented in Matlab software environment. Chosen method for solving the task uses the Condat algorithm or Generic proximal algorithm for convex optimization and solves minimization of sum of three convex functions. The result of the thesis is five versions of algorithm and three of them have implemented psychoacoustic model for results improvement. For each version has been found optimal setting of parameters. The restoration quality results are evaluated using objective measurements like SDR and PEMO-Q and also using subjective listening test.