Multiple classifier system has shown to be an effective technique for classification. The success of multiple classifiers does not entirely depend on the base classifiers and/or the aggregation technique. Other parameters, such as training data, feature attributes, and correlation among the base classifiers may also contribute to the success of multiple classifiers. In addition, interaction of these parameters with each other may have an impact on multiple classifiers performance. In the present study, we intended to examine some of these interactions and investigate further the effects of these interactions on the performance of classifier ensembles. The proposed research introduces a different direction in the field of multiple classifiers systems. We attempt to understand and compare ensemble methods from the cooperation perspective. In this thesis, we narrowed down our focus on cooperation at training level. We first developed measures to estimate the degree and type of cooperation among training data partitions. These evaluation measures enabled us to evaluate the diversity and correlation among a set of disjoint and overlapped partitions. With the aid of properly selected measures and training information, we proposed two new data partitioning approaches: Cluster, De-cluster, and Selection (CDS) and Cooperative Cluster, De-cluster, and Selection (CO-CDS). In the end, a comprehensive comparative study was conducted where we compared our proposed training approaches with several other approaches in terms of robustness of their usage, resultant classification accuracy and classification stability. Experimental assessment of CDS and CO-CDS training approaches validates their robustness as compared to other training approaches. In addition, this study suggests that: 1) cooperation is generally beneficial and 2) classifier ensembles that cooperate through sharing information have higher generalization ability compared to the ones that do not share training information.

Multiple classifier system has shown to be an effective technique for classification. The success of multiple classifiers does not entirely depend on the base classifiers and/or the aggregation technique. Other parameters, such as training data, feature attributes, and correlation among the base classifiers may also contribute to the success of multiple classifiers. In addition, interaction of these parameters with each other may have an impact on multiple classifiers performance. In the present study, we intended to examine some of these interactions and investigate further the effects of these interactions on the performance of classifier ensembles. The proposed research introduces a different direction in the field of multiple classifiers systems. We attempt to understand and compare ensemble methods from the cooperation perspective. In this thesis, we narrowed down our focus on cooperation at training level. We first developed measures to estimate the degree and type of cooperation among training data partitions. These evaluation measures enabled us to evaluate the diversity and correlation among a set of disjoint and overlapped partitions. With the aid of properly selected measures and training information, we proposed two new data partitioning approaches: Cluster, De-cluster, and Selection (CDS) and Cooperative Cluster, De-cluster, and Selection (CO-CDS). In the end, a comprehensive comparative study was conducted where we compared our proposed training approaches with several other approaches in terms of robustness of their usage, resultant classification accuracy and classification stability. Experimental assessment of CDS and CO-CDS training approaches validates their robustness as compared to other training approaches. In addition, this study suggests that: 1) cooperation is generally beneficial and 2) classifier ensembles that cooperate through sharing information have higher generalization ability compared to the ones that do not share training information.

The traditional education system in the United States developed a goal of sending a high number of students to post secondary schools. The expectation of going to college starts at elementary school and continues through high school and is pushed by the school district’s hierarchy. Schools annually publish numbers through the Department of Education at state and national levels of students moving to college. College progression has become such a priority that many students feel they have failed if not college bound. Students not considered academic often move toward vocational education with the intent of post-secondary education to support a chosen career field. Students with lower grades and abilities are routed to vocational classes utilizing a shop model. The education system identifies and educates students at expected age and grade level requirements. High stakes testing has been established to determine if students meet those educational requirements considered necessary for success. Students not meeting established math, English and reading requirements are given remediation to bring academic skills to appropriate age and grade levels. My thesis described a contextual reading program for building reading skills. Content area material is used to teach academic skills in writing, reading and math to levels supporting high stakes testing needed for success. Contextual reading has improved reading competencies for vocational education students in the CATER program at Kathleen High School in the Polk County School System as well as in the other statewide settings. This has the potential to improve career and technical education (CTE) in other systems. A contextual approach also has potential to improve teaching and learning in socalled academic areas, and improving scores in high stakes tests. Recommendations for further study include following this line of action for younger and adult students

This thesis aims to evaluate the current state of the art for unconditional text generation and compare established models with novel approaches in the task of generating texts, after being trained on texts written by political parties from the Swedish Riksdag. First, the progression of language modeling from n-gram models and statistical models to neural network models is presented. This is followed by theoretical arguments for the development of adversarial training methods,where a generator neural network tries to fool a discriminator network, trained to distinguish between real and generated sentences. One of the methods in the research frontier diverges from the adversarial idea and instead uses cooperative training, where a mediator network is trained instead of a discriminator. The mediator is then used to estimate a symmetric divergence measure between the true distribution and the generator’s distribution, which is to be minimized in training. A set of experiments evaluates the performance of cooperative training and adversarial training, and finds that they both have advantages and disadvantages. In the experiments, the adversarial training increases the quality of generated texts, while the cooperative training increases the diversity. The findings are in line with the theoretical expectation. / Denna uppsats utvärderar några nyligen föreslagna metoder för obetingad textgenerering, baserade på s.k. “Generative Adversarial Networks” (GANs). Den jämför etablerade modeller med nya metoder för att generera text, efter att ha tränats på texter från de svenska Riksdagspartierna. Utvecklingen av språkmodellering från n-gram-modeller och statistiska modeller till modeller av neurala nätverk presenteras. Detta följs upp av teoretiska argument för utvecklingen av GANs, för vilka ett generatornätverk försöker överlista ett diskriminatornätverk, som tränas skilja mellan riktiga och genererade meningar. En av de senaste metoderna avviker från detta angreppssätt och introducerar istället kooperativ träning, där ett mediatornätverk tränas istället för en diskriminator. Mediatorn används sedan till att uppskatta ett symmetriskt divergensmått mellan den sanna distributionen och generatorns distribution, vilket träningen syftar till att minimera. En serie experiment utvärderar hur GANs och kooperativ träning presterar i förhållande till varandra, och finner att de båda har för- och nackdelar. I experimenten ökar GANs kvaliteten på texterna som genereras, medan kooperativ träning ökar mångfalden. Resultaten motsvarar vad som kan förväntas teoretiskt.

L'enquête empirique menée dans le cadre de cette thèse s'inscrivait dans la perspective de la Loi de Refondation « Pour une école de la bienveillance ». Il s'agissait de contribuer à un programme de recherche de l'Institut Français de l'Éducation (IFÉ), nommé Lieux d'Éducation Associés (LéA). L'objet de cette étude (2014-2017) s'appelait « Innover pour une école bienveillante ». Le terrain d'enquête est celui de l'École Freinet, école historique d'Élise et Célestin Freinet à Vence dans les Alpes-Maritimes (France). En séjournant régulièrement in situ, les pratiques effectives lors de situations d'écriture, propres à l'institution du texte libre, ont été objectivées. Un tel travail d'objectivation donne à voir l'action conjointe des professeurs et des formatrices. Ces pratiques coopératives de formation ont été analysées en s'appuyant notamment sur le cadre de la Théorie de l'Action Conjointe en Didactique (TACD).Cette thèse se propose d'examiner une série de questions : Quel modèle est-il possible de proposer du type de formation pratiqué à l’École Freinet ? En quoi ce modèle, que l'on peut qualifier de clinique, est-il pertinent lorsqu'il s’agit d’une filiation en pédagogie ? Quelles en sont les forces, les faiblesses et les limites ? Ces questions s'articulent à une autre série : En quoi peut-on parler d'une bienveillance professionnelle du professeur ? En quelles autres vertus professionnelles pourrait-elle s’exprimer ? / The empirical investigation carried out within the framework of this thesis was part of the perspective of the Law of Refoundation "For a school of benevolence". It was to contribute to a research program of the French Institute of Education (IFE), named Places of Education Associates (LEA). The purpose of this study (2014-2017) was "Innovating for a caring school of benevolence". The field of investigation is that of the Freinet School, the historical school of Élise and Célestin Freinet in Vence in the Alpes-Maritimes (France). By staying regularly in situ, the actual practices during writing situations, specific to the institution of the free text, have been objectified. Such a work of objectification pinpoints the joint action of the teachers. These cooperative training practices were analyzed based in particular on the framework of the Theory of Joint Action in Didactics (TACD). This thesis proposes to examine a series of questions: What model is it possible to propose from the type of training practiced at Freinet School? How is this model, which can be described as clinical, relevant when it comes to pedagogical filiation? What are the strengths, weaknesses and limitations? These questions are linked to another series: How can one speak of the teacher's professional benevolence? In what other professional virtues could it find its expression?

This qualitative research investigated the implementation problems of apprenticeship training program in the Addis Ababa TVET colleges and enterprises. The research looked into the training program conducted by TVET Colleges and enterprises. Relevant literature on the nature of apprenticeship training and the factors that could affect its implementation were also reviewed. The research applied the theoretical statement of Bandura’s (1977) theory of observation /imitation/, Vygotsky’s theory of ZPD, contextual or the situated learning (Lave and Wenger, 1991), the constructivist view of learning and career theory as the main theoretical frameworks to describe the apprenticeship training process. The theories, however, do not mention the factors that can hamper the implementation of apprenticeship training in a specific social context. The theory of observation was preferred since it can describe the apprentice’s effort to master the skill of an occupation by observing the craftsman at the enterprises. The situated learning was preferred as it gives value to practice as a condition for the effectiveness of learning. Similarly, the research considered Vygotsky’s theory of the ZPD. The application of ZPD indicates the difference between what the apprentice can accomplish independently and what he/she can do with the close assistance of supervisor on a given task. Holland’s career theory (as cited in De lary, Duncan & Swarth, 2006) is also considered to describe the apprenticeship training in relation with an occupation. The current research has enabled to consider context specific conditions by focusing on the investigation of the causes of the implementation challenges of the apprenticeship training. The researcher collected data from Addis Ababa TVET Agency experts, deans of one private and another public TVET Colleges, college apprenticeship training coordinators, trainers, trainees, supervisors and enterprises by employing unstructured and structured interviews, observation and Focus Group Discussion tools. The analysis was made by making the data pass through three successive steps: data reduction, data display and interpretation, and conclusion. The findings showed that the implementation of the apprenticeship training program had challenges that could be shared among its actors; Trade Unions had no involvement in the implementation of the apprenticeship training program; the selected colleges and enterprises addressed their challenges by using limited strategies and without regular and systematic way; the presence of some less motivating factors for the participation of the apprentices in the apprenticeship training was reported; the presence of conditions that could adversely affect the apprentices’ acquisition of occupational skills was reported; the provision of apprenticeship training at the enterprises lack, either facilities or training services, in order to arrange the apprenticeship training program for TVET college trainees; and there were some indications for the presence of some weak professional qualities of supervisors that have implications for the apprentices’ training. Eventually, it was concluded that the implementation of apprenticeship training program in the Addis Ababa TVET Colleges and enterprises had challenges that require the consideration of both institutional and human conditions. / Curriculum and Instructional Studies / D. Ed. (Curriculum Studies)

This qualitative research investigated the implementation problems of apprenticeship training program in the Addis Ababa TVET colleges and enterprises. The research looked into the training program conducted by TVET Colleges and enterprises. Relevant literature on the nature of apprenticeship training and the factors that could affect its implementation were also reviewed. The research applied the theoretical statement of Bandura’s (1977) theory of observation /imitation/, Vygotsky’s theory of ZPD, contextual or the situated learning (Lave and Wenger, 1991), the constructivist view of learning and career theory as the main theoretical frameworks to describe the apprenticeship training process. The theories, however, do not mention the factors that can hamper the implementation of apprenticeship training in a specific social context. The theory of observation was preferred since it can describe the apprentice’s effort to master the skill of an occupation by observing the craftsman at the enterprises. The situated learning was preferred as it gives value to practice as a condition for the effectiveness of learning. Similarly, the research considered Vygotsky’s theory of the ZPD. The application of ZPD indicates the difference between what the apprentice can accomplish independently and what he/she can do with the close assistance of supervisor on a given task. Holland’s career theory (as cited in De lary, Duncan & Swarth, 2006) is also considered to describe the apprenticeship training in relation with an occupation. The current research has enabled to consider context specific conditions by focusing on the investigation of the causes of the implementation challenges of the apprenticeship training. The researcher collected data from Addis Ababa TVET Agency experts, deans of one private and another public TVET Colleges, college apprenticeship training coordinators, trainers, trainees, supervisors and enterprises by employing unstructured and structured interviews, observation and Focus Group Discussion tools. The analysis was made by making the data pass through three successive steps: data reduction, data display and interpretation, and conclusion. The findings showed that the implementation of the apprenticeship training program had challenges that could be shared among its actors; Trade Unions had no involvement in the implementation of the apprenticeship training program; the selected colleges and enterprises addressed their challenges by using limited strategies and without regular and systematic way; the presence of some less motivating factors for the participation of the apprentices in the apprenticeship training was reported; the presence of conditions that could adversely affect the apprentices’ acquisition of occupational skills was reported; the provision of apprenticeship training at the enterprises lack, either facilities or training services, in order to arrange the apprenticeship training program for TVET college trainees; and there were some indications for the presence of some weak professional qualities of supervisors that have implications for the apprentices’ training. Eventually, it was concluded that the implementation of apprenticeship training program in the Addis Ababa TVET Colleges and enterprises had challenges that require the consideration of both institutional and human conditions. / Curriculum and Instructional Studies / D. Ed. (Curriculum Studies)