Made available in DSpace on 2014-08-20T13:47:53Z (GMT). No. of bitstreams: 1 Eliana Rita Mariotto Colpo_Dissertacao.pdf: 1649520 bytes, checksum: b6c6d2f9cf870a2edd0085dffd5d34df (MD5) Previous issue date: 2012-11-05 / The present piece of work was developed within the Research Group on Imaginary, Education and Memory (GEPIEM), a Master s degree Graduation Program in Education, in the field of Writing Culture: languages and learning. It has the analysis of the literary work "The Catcher in the Rye" by J. D. Sallinger as its methodological direction, with the aim of seeking convergence from this "novel of formation" (Bildungsroman) for reflections in education .In particular, to look for its contributions, in the sense of the composition between studies and the imaginary (self) human formation. The analyzed work is considered as a powerful symbolic and ideological device for being able to transit between a teenager and his conflicts, sufferings and individual drives in affective, social, and cultural contexts and the wider reach of the society of the time - mid-XX century. The narrative of the main character and his relationships with family, school peers and friends has a timeless dimension, ie, that can belong to any time, and this, therefore, makes the work so contemporary. From the reading of the work we extract some symbolic cores grouped into four "mythemes" able to show the wealth and power of the "novel of formation" for (self) human training. The first mytheme: a human can (self) be formed between conflict and madness in search of the treasure inside. The second mytheme: a (self) formation can take place on a trip between escape and confrontation. The third mytheme: a (self) human formation occurs by initiation and transmigration. In the last and fourth mytheme: the fall into the abyss as act of living and (self) forming. Such mythemes were analyzed in the light of studies of imaginary, especially in its symbolic aspects, in order to relate and appreciate the importance of reading the "novel of formation" and the writings of other ways to raise life to the education of the human being. In a perception that values the felt and lived things, as well as the known and proven ones. / O presente trabalho de pesquisa nasce no interior do Grupo de Estudos e Pesquisas sobre Imaginário, Educação e Memória (GEPIEM), no Programa de Pós-graduação em Educação, nível mestrado, na linha de Cultura Escrita: linguagens e aprendizagem. Tendo como norte metodológico a análise da obra literária O Apa-nhador no Campo de Centeio , de J. D. Sallinger, com o intuito de buscar conver-gências deste romance de formação (bildungsroman) para pensar a Educação. Em especial, buscar as contribuições desta a partir da tessitura entre os estudos do imaginário e a (auto) formação humana. A obra analisada constitui-se como um po-tente ideologema por ser capaz de fazer o trânsito entre um adolescente e seus conflitos, sofrimentos e pulsões individuais em seu contexto afetivo, social, cultural e o âmbito mais geral da sociedade da época - metade do século XX. A narrativa do protagonista e suas relações com a família, escola, colegas e amigos possui uma dimensão atemporal, ou seja, que pode pertencer a qualquer tempo, tornando a obra, portanto, contemporânea. A partir da leitura da obra extraímos alguns núcleos simbólicos agrupados em quatro mitemas capazes de mostrar a riqueza e a potên-cia do romance de formação para a (auto)formação humana. O primeiro mitema: o ser humano pode (auto) formar-se entre o conflito e a loucura na busca do te-souro interior. O segundo mitema: a (auto) formação pode dar-se na viagem en-tre a evasão e o confronto. O terceiro mitema: a (auto) formação humana ocorre pela iniciação e a transmigração. No último e quarto mitema: a queda no abismo como ato de viver e (auto) formar-se. Tais mitemas foram analisados à luz dos estudos do imaginário, em especial em seus aspectos simbólicos, no sentido de re-lacionar e valorizar a importância das leituras do romance de formação e das escri-tas de vida para suscitar outros sentidos à educação do ser humano. Uma percep-ção que valorize o sentido e o vivido, tanto quanto o sabido e o comprovado.

The self-training movement of a teacher creating education workshop strategies (CORREA, 2000 & 2006) potentiated a research development with an interest of map the acquired knowledge during the immersion on deaf science education, on Escola Estadual de Educação Especial Doutor Reinaldo Fernando Coser, in Santa Maria/RS. The focus of this study is to report through diary and notes the reflection movements of a teacher in front of the question how should a science concept be shown if there is no signal in LIBRAS? The notes were produced during a science teacher training and on workshop realized through the extension project at school during the Chemistry Graduation. These writings shows an estrangement on teachers formation that is comprehend with the subjectivity machine of a teacher body (CORREA & PREVE, 2011) which the effect is exhaustion of teacher capacity to create different strategies of study to answer tests . In this cartography (ROLNIK, 2007) (GUATARRI & ROLNIK, 2011) the science education with deaf is comprehend such as a study territory (FREIRE, 1988) in know process with will (STIRNER, 2001). And giving a language (ROLNIK, 2007) to these sensations present on the diary notes grow such as a necessity of speak what still has not been said to Natural Science teachers. Giving space to visualization of these notes allowing find the acquired knowledge from bifurcation realized by the teacher in front of his questions which compose his self-training process. / O movimento de autoformação de um educador na elaboração de estratégias de educação em oficinas (CORREA, 2000 & 2006) potencializou o desenvolvimento de uma pesquisa interessada em mapear os saberes adquiridos durante a sua imersão na educação em Ciências com surdos, na Escola Estadual de Educação Especial Doutor Reinaldo Fernando Coser, em Santa Maria/RS. O foco desse estudo é relatar através de diários e anotações os movimentos de reflexão do educador a partir da questão como apresentar um conceito em Ciências se não há sinais em LIBRAS? . As anotações foram produzidas durante os estágios em Ciências e na produção de oficinas realizadas através de projetos de extensão na escola durante a graduação em Licenciatura em Química. Tais escritos apontam para o estranhamento da formação de educadores a qual é compreendida com uma maquinaria de subjetivação do corpo do educador (CORREA & PREVE, 2011) cujo efeito é esgotamento da capacidade do educador em criar estratégias diferenciadas de ensinar a fim de responder avaliações . Nessa cartografia (ROLNIK, 2007) (GUATARRI & ROLNIK, 2011) a educação em Ciências com surdos é compreendida como um território de estudo (FREIRE, 1988) em processos de conhecer com vontade (STIRNER, 2001). E dar linguagem (ROLNIK, 2007) a essas sensações presentes nas anotações dos diários emerge como uma necessidade de se falar o que ainda não foi dito aos educadores em Ciências Naturais. Dar espaço para a visualização dessas anotações permite encontrar os saberes adquiridos a partir das bifurcações realizadas pelo educador diante das suas questões as quais compõe o seu processo de autoformação.

Semi-supervised self-learning algorithms have been shown to improve classifier accuracy under a variety of conditions. In this thesis, semi-supervised self-learning using ensembles of random forests and fuzzy c-means clustering similarity was applied to three data sets to show where improvement is possible over random forests alone. Two of the data sets are emulations of large simulations in which the data may be distributed. Additionally, the ratio of majority to minority class examples in the training set was altered to examine the effect of training set bias on performance when applying the semi-supervised algorithm.

One of the most important business elements in primary and secondary industries is monitoring their equipment. Understanding the usage of heavy logistics machinery can help in realizing the potential of these machinery and improving them. With the purpose of monitoring and quantifying machine usage, Machine Activity Recognition (MAR) problems can be solved with machine learning techniques. In this project, We propose a method of augmented Self-training which collaborates Self-training and data augmentation to solve forklift trucks' MAR problem on Controller Area Network (CAN bus) data. Compared to the standard Self-training method, the augmented Self-training performs data augmentation on pseudo-labeled data to inject noise and to improve model generalization. The best student model of the augmented Self-training achieves 71.8% balanced accuracy (BA) with improvement of 3.0% from applying Supervised Learning solely (68.8% BA). In addition, Matthews Correlation Coefficient (MCC) for the augmented Self-training's best student model reaches 0.658 with an increment of 0.031, compare to an MCC of 0.627 by only applying Supervised Learning. The augmented Self-training improves model performance.

This thesis proposes a novel approach for efficiently estimating gaze points in dual camera transportation datasets. Traditional methods for gaze point estimation are dependent on large amounts of labeled data, which can be both expensive and time-consuming to collect. Additionally, alignment and calibration of the two camera views present significant challenges. To overcome these limitations, this thesis investigates the use of self-learning techniques such as semi-supervised learning and self-training, which can reduce the need for labeled data while maintaining high accuracy. The proposed method is evaluated on the DGAZE dataset and achieves a 57.2\% improvement in performance compared to the previous methods. This approach can prove to be a valuable tool for studying visual attention in transportation research, leading to more cost-effective and efficient research in this field. / Master of Science / This thesis presents a new method for efficiently estimating the gaze point of drivers while driving, which is crucial for understanding driver behavior and improving transportation safety. Traditional methods require a lot of labeled data, which can be time-consuming and expensive to obtain. This thesis proposes a self-learning approach that can learn from both labeled and unlabeled data, reducing the need for labeled data while maintaining high accuracy. By training the model on labeled data and using its own estimations on unlabeled data to improve its performance, the proposed approach can adapt to new scenarios and improve its accuracy over time. The proposed method is evaluated on the DGAZE dataset and achieves a 57.2\% improvement in performance compared to the previous methods. Overall, this approach offers a more efficient and cost-effective solution that can potentially help improve transportation safety by providing a better understanding of driver behavior. This approach can prove to be a valuable tool for studying visual attention in transportation research, leading to more cost-effective and efficient research in this field.

La compréhension automatique de la parole est au confluent des deux grands domaines que sont la reconnaissance automatique de la parole et l'apprentissage automatique. Un des problèmes majeurs dans ce domaine est l'obtention d'un corpus de données conséquent afin d'obtenir des modèles statistiques performants. Les corpus de parole pour entraîner des modèles de compréhension nécessitent une intervention humaine importante, notamment dans les tâches de transcription et d'annotation sémantique. Leur coût de production est élevé et c'est la raison pour laquelle ils sont disponibles en quantité limitée.Cette thèse vise principalement à réduire ce besoin d'intervention humaine de deux façons : d'une part en réduisant la quantité de corpus annoté nécessaire à l'obtention d'un modèle grâce à des techniques d'apprentissage semi-supervisé (Self-Training, Co-Training et Active-Learning) ; et d'autre part en tirant parti des réponses de l'utilisateur du système pour améliorer le modèle de compréhension.Ce dernier point touche à un second problème rencontré par les systèmes de compréhension automatique de la parole et adressé par cette thèse : le besoin d'adapter régulièrement leurs modèles aux variations de comportement des utilisateurs ou aux modifications de l'offre de services du système / Two wide research fields named Speech Recognition and Machine Learning meet with the Automatic Speech Language Understanding. One of the main problems in this domain is to obtain a sufficient corpus to train an efficient statistical model. Such speech corpora need a lot of human involvement to transcript and semantically annotate them. Their production cost is therefore quite high and they are difficultly available.This thesis mainly aims at reducing the need of human intervention in two ways: firstly, reducing the amount of corpus needed to build a model thanks to some semi-supervised learning methods (Self-Training, Co-Training and Active-Learning); And lastly, using the answers of the system end-user to improve the comprehension model.This last point addresses another problem related to automatic speech understanding systems: the need to adapt their models to the fluctuation of end-user habits or to the modification of the services list offered by the system

Algoritmos de aprendizado semissupervisionado aprendem a partir de uma combinação de dados rotulados e não rotulados. Assim, eles podem ser aplicados em domínios em que poucos exemplos rotulados e uma vasta quantidade de exemplos não rotulados estão disponíveis. Além disso, os algoritmos semissupervisionados podem atingir um desempenho superior aos algoritmos supervisionados treinados nos mesmos poucos exemplos rotulados. Uma poderosa abordagem ao aprendizado semissupervisionado, denominada aprendizado multidescrição, pode ser usada sempre que os exemplos de treinamento são descritos por dois ou mais conjuntos de atributos disjuntos. A classificação de textos é um domínio de aplicação no qual algoritmos semissupervisionados vêm obtendo sucesso. No entanto, o aprendizado semissupervisionado multidescrição ainda não foi bem explorado nesse domínio dadas as diversas maneiras possíveis de se descrever bases de textos. O objetivo neste trabalho é analisar o desempenho de algoritmos semissupervisionados multidescrição na classificação de textos, usando unigramas e bigramas para compor duas descrições distintas de documentos textuais. Assim, é considerado inicialmente o difundido algoritmo multidescrição CO-TRAINING, para o qual são propostas modificações a fim de se tratar o problema dos pontos de contenção. É também proposto o algoritmo COAL, o qual pode melhorar ainda mais o algoritmo CO-TRAINING pela incorporação de aprendizado ativo como uma maneira de tratar pontos de contenção. Uma ampla avaliação experimental desses algoritmos foi conduzida em bases de textos reais. Os resultados mostram que o algoritmo COAL, usando unigramas como uma descrição das bases textuais e bigramas como uma outra descrição, atinge um desempenho significativamente melhor que um algoritmo semissupervisionado monodescrição. Levando em consideração os bons resultados obtidos por COAL, conclui-se que o uso de unigramas e bigramas como duas descrições distintas de bases de textos pode ser bastante compensador / Semi-supervised learning algorithms learn from a combination of both labeled and unlabeled data. Thus, they can be applied in domains where few labeled examples and a vast amount of unlabeled examples are available. Furthermore, semi-supervised learning algorithms may achieve a better performance than supervised learning algorithms trained on the same few labeled examples. A powerful approach to semi-supervised learning, called multi-view learning, can be used whenever the training examples are described by two or more disjoint sets of attributes. Text classification is a domain in which semi-supervised learning algorithms have shown some success. However, multi-view semi-supervised learning has not yet been well explored in this domain despite the possibility of describing textual documents in a myriad of ways. The aim of this work is to analyze the effectiveness of multi-view semi-supervised learning in text classification using unigrams and bigrams as two distinct descriptions of text documents. To this end, we initially consider the widely adopted CO-TRAINING multi-view algorithm and propose some modifications to it in order to deal with the problem of contention points. We also propose the COAL algorithm, which further improves CO-TRAINING by incorporating active learning as a way of dealing with contention points. A thorough experimental evaluation of these algorithms was conducted on real text data sets. The results show that the COAL algorithm, using unigrams as one description of text documents and bigrams as another description, achieves significantly better performance than a single-view semi-supervised algorithm. Taking into account the good results obtained by COAL, we conclude that the use of unigrams and bigrams as two distinct descriptions of text documents can be very effective

Made available in DSpace on 2014-12-17T14:36:27Z (GMT). No. of bitstreams: 1 SandraLP_TESE.pdf: 2320249 bytes, checksum: 3f1b3e5febdd6473ad8eea2fce8ee234 (MD5) Previous issue date: 2012-08-03 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / The object of study of this thesis is the use of (self)training workshops as a fundamental process for the constitution of the teaching subject in mathematics education. The central purposes of the study were to describe and analyze a learning process of mathematics teachers supported by the training-research methodology, which procedures have been affected with the practice of (self)training workshops as a way of collaborating to the constitution of the teaching subject in Mathematics Education. The survey was conducted with a group of teachers in the city of Nova Cruz, Rio Grande do Norte through a process of continued education realized in the training workshops having as main goal the realization of the group s (self)training sessions in order to lead participants to the extent of their autonomy in their personal and professional transformations. The results obtained in the formative processes have shown the need to develop activities of mathematics teaching as a contribution to overcome the conceptual difficulties of the teachers, apart from their (self)reflections about themselves and the educational processes in which they belong. The results raised some propositions about (self)training workshops that may be incurred in practices to be included in the curriculum frameworks or materialize as a strategy of pedagogical work in training courses for teachers of mathematics. Also, they can constitute an administrative and educational activity to be instituted in the public schools of Basic Education / O objeto de estudo que origina esta tese ? o uso de ateli?s (auto)formativos como um processo fundamental para a constitui??o do sujeito docente da Educa??o Matem?tica. As finalidades centrais do trabalho foram descrever e analisar um processo formativo de professores de Matem?tica apoiado na metodologia da pesquisa-forma??o, cujos procedimentos se efetivaram com a pr?tica de ateli?s (auto)formativos como uma forma de contribuir para a constitui??o do sujeito docente na Educa??o Matem?tica. A pesquisa foi realizada com um grupo de professores de matem?tica do munic?pio de Nova Cruz, Rio Grande do Norte por meio de um processo de forma??o continuada concretizado nos ateli?s formativos tendo como meta principal a realiza??o de sess?es (auto)formativas do grupo com vistas a levar os participantes ao alcance de sua autonomia nas suas transforma??es pessoais e profissionais. Os resultados obtidos nos processos formativos mostraram a necessidade de se desenvolver atividades de ensino de Matem?tica como contribui??o para superar as dificuldades conceituais dos professores, al?m das suas (auto)reflex?es sobre si e sobre os processos educativos aos quais est?o inseridos. Os resultados suscitaram algumas proposi??es acerca de ateli?s (auto)formativos que poder?o se constituir em pr?ticas a serem inclu?das nas estruturas curriculares ou materializar-se como estrat?gia de trabalho pedag?gico nos cursos de forma??o de professores de Matem?tica. Al?m disso, podem se constituir em uma atividade administrativo-pedag?gica a ser institu?da nas escolas p?blicas de Educa??o B?sica

Algoritmos de aprendizado semissupervisionado aprendem a partir de uma combinação de dados rotulados e não rotulados. Assim, eles podem ser aplicados em domínios em que poucos exemplos rotulados e uma vasta quantidade de exemplos não rotulados estão disponíveis. Além disso, os algoritmos semissupervisionados podem atingir um desempenho superior aos algoritmos supervisionados treinados nos mesmos poucos exemplos rotulados. Uma poderosa abordagem ao aprendizado semissupervisionado, denominada aprendizado multidescrição, pode ser usada sempre que os exemplos de treinamento são descritos por dois ou mais conjuntos de atributos disjuntos. A classificação de textos é um domínio de aplicação no qual algoritmos semissupervisionados vêm obtendo sucesso. No entanto, o aprendizado semissupervisionado multidescrição ainda não foi bem explorado nesse domínio dadas as diversas maneiras possíveis de se descrever bases de textos. O objetivo neste trabalho é analisar o desempenho de algoritmos semissupervisionados multidescrição na classificação de textos, usando unigramas e bigramas para compor duas descrições distintas de documentos textuais. Assim, é considerado inicialmente o difundido algoritmo multidescrição CO-TRAINING, para o qual são propostas modificações a fim de se tratar o problema dos pontos de contenção. É também proposto o algoritmo COAL, o qual pode melhorar ainda mais o algoritmo CO-TRAINING pela incorporação de aprendizado ativo como uma maneira de tratar pontos de contenção. Uma ampla avaliação experimental desses algoritmos foi conduzida em bases de textos reais. Os resultados mostram que o algoritmo COAL, usando unigramas como uma descrição das bases textuais e bigramas como uma outra descrição, atinge um desempenho significativamente melhor que um algoritmo semissupervisionado monodescrição. Levando em consideração os bons resultados obtidos por COAL, conclui-se que o uso de unigramas e bigramas como duas descrições distintas de bases de textos pode ser bastante compensador / Semi-supervised learning algorithms learn from a combination of both labeled and unlabeled data. Thus, they can be applied in domains where few labeled examples and a vast amount of unlabeled examples are available. Furthermore, semi-supervised learning algorithms may achieve a better performance than supervised learning algorithms trained on the same few labeled examples. A powerful approach to semi-supervised learning, called multi-view learning, can be used whenever the training examples are described by two or more disjoint sets of attributes. Text classification is a domain in which semi-supervised learning algorithms have shown some success. However, multi-view semi-supervised learning has not yet been well explored in this domain despite the possibility of describing textual documents in a myriad of ways. The aim of this work is to analyze the effectiveness of multi-view semi-supervised learning in text classification using unigrams and bigrams as two distinct descriptions of text documents. To this end, we initially consider the widely adopted CO-TRAINING multi-view algorithm and propose some modifications to it in order to deal with the problem of contention points. We also propose the COAL algorithm, which further improves CO-TRAINING by incorporating active learning as a way of dealing with contention points. A thorough experimental evaluation of these algorithms was conducted on real text data sets. The results show that the COAL algorithm, using unigrams as one description of text documents and bigrams as another description, achieves significantly better performance than a single-view semi-supervised algorithm. Taking into account the good results obtained by COAL, we conclude that the use of unigrams and bigrams as two distinct descriptions of text documents can be very effective

La recherche présentée s’intéresse à l’implication des Technologies de l’Information et de la Communication (TIC) dans le développement de l’autoformation des professeurs des écoles stagiaires. Elle s’inscrit dans un double contexte. D’une part, celui de la mise en place de la réforme de la formation initiale des enseignants – la masterisation –, ayant eu cours, en France, de 2010 à 2013. D’autre part, celui des problématiques contemporaines liées à la question des usages du numérique à des fins d’autoformation. Dans cette perspective, ce travail s’appuie sur les données d’entretiens individuels, réalisés auprès de professeurs des écoles stagiaires issus de la masterisation, afin d’explorer leur rapport au métier, au(x) savoir(s) et aux technologies de l’information et de la communication. Il vise à mieux comprendre ce que représente, pour eux, l’autoformation aux et par les TIC et à saisir ce que cela soulève d’enjeux, en termes de construction et de développement professionnel. Il en ressort qu’à la complexité de cette période tout à fait particulière d’entrée dans le métier, s’ajoute aujourd’hui pour les enseignants débutants, celle d’enseigner mais également de se former et de s’autoformer dans un « monde numérique ». Il semble que les potentialités qu’offrent les technologies ne font que renforcer l’importance fondamentale d’une formation initiale solide, passant avant tout par un renforcement des interactions humaines, par un accompagnement éclairé et éclairant des enseignants débutants, leur permettant de développer au plus tôt leurs compétences professionnelles et d’être à même de les faire évoluer tout au long de leur carrière. / The present research treats of the involvement of Information and Communication Technologies in the development of the self-training of to-be primary school teachers. The context in which this research settles is double. Firstly, that of the setting up of the teachers’ training reform – masterisation – which took place in France from 2010 to 2013. Secondly, that of the contemporary problems related to the use of digital technology for self-training. The following work is based on data which was gathered during individual interviews with primary school teacher trainees who originated from masterisation, in order to explore their views on the profession, on the knowledge as well as on Information and Communication Technologies. Its goal is the better understanding of what self-training via ITCs represents for them, and also to bring into light the stakes engendered, in terms of career building and development. What stands out from this, is that the complexity of this rather peculiar pre-teaching phase is doubled by that of teaching and of self-training in a “digital world”. It appears that the infinite possibilities which are offered by these technologies only strengthen the need for a very solid initial training. This implies the reinforcement of human interaction through an enlightened and enlightening guidance for beginning teachers, enabling them to develop all the sooner their professional competence and to continue to make them evolve all along their career.