From the author’s work as an educational psychologist in a Scottish local authority and her initial literature review, concerns were raised by her colleagues and local education authority about delaying school entry of children with additional support needs. This study aimed to: explore how the decision making process for delaying school entry operated; develop and trial a methodology to capture the children’s perspective and explore the experiences of the children and their families in nursery and during their first year of school. A qualitative case study approach was used to explore the decision making process for six children and their families and the experiences of five children and their families. Techniques from the mosaic methodology (Clark and Moss, 2001) were adapted to capture the children’s voices. The study revealed that participants held different models of school readiness, in line with other international research, and this influenced their decision to retain. Participants reported a range of positive and negative outcomes of delayed school entry and identified factors that had been supportive in managing the children’s transitions. Children’s perspectives were captured and often offered a unique insight into the children’s views. At the end of the children’s first year in primary school four parents were still happy with their decision to delay school entry but one parent regretted her decision. In conclusion it is suggested that a more interactionist (Meisels, 1998) approach to school readiness should be taken when supporting the transition from nursery to school. A moderated system should be developed to support the decision making process for delaying school entry, with an inbuilt process to follow up longer-term outcomes for the children and families concerned. The methodology developed for gathering children’s views offered a way for children with additional support needs to express their own views and take an active role at this important transition point. There is potential to develop this methodology further to ensure that children’s voices are heard in this first universal educational transition.

Rose-scented geranium (Pelargonium capitatum x P. radens) belongs to the family Geraniaceae and it is a multi-harvest, high value, commercially important essential oil yielding aromatic plant. The essential oil extracted from the herbage of the plant is widely used in the fragrance and cosmetics industry and scenting of soaps. The essential oil is extracted by steam distillation. South Africa is now producing significant quantities of geranium oil. However, previous experience by this research group showed that seedlings often take long to establish, resulting in high death rates and sometimes poor growth after establishment. Stunted growth and yellowing of leaves was also observed in some cases. Poor vegetative growth causes low herbage yield and, consequently, low total essential oil production per hectare. Poor growth is believed to be due to a combination of factors, including nutrient deficiencies and acidic soil conditions and has not been studied under South Africa condition. Production of the crop is also reported to respond differently to rate and source of nutrition in various agro-ecological regions of the world. Therefore, field trials were conducted at the Hatfield Experimental Farm, University of Pretoria, in order to investigate how the plant responds to agronomic practices, such as source and amount of nitrogen, time of N fertilizer application, season of N fertilization and liming. Response of rose-scented geranium to source and amount of N showed that, at the first harvest (summer/autumn), there was no significant effect of conventional N on fresh herbage and oil yield, probably due to leaching of N by rainfall. However, organic N at 100 kg•ha-1 increased fresh herbage and oil yields by 58% and 48% over the control, respectively. In the second harvest (spring/summer), fresh herbage yield increased by 46% (conventional N) and 60% (organic N) at 100 kg•ha-1 compared to the control. Compared to the control, 100 kg•ha-1 conventional and organic N also increased essential oil yields by 94% and 129%, respectively. For both N sources nitrogen use efficiency (NUE) and LAI decreased with an increase in N level, and organic N gave highest essential oil production efficiency and LAI. Essential oil content (% fresh mass basis) also varied between the harvests, being greater in the second harvest (September to December 2005; spring/summer) than the first harvest (February to May 2005; summer/autumn). This was due to environmental variations that occurred between the harvesting periods. N level and source were found to have no noticeable effect on essential oil composition. This study revealed that rose-scented geranium produced higher fresh herbage and essential oil yield when organic fertilizer was used as a source of N. Nitrogen management in terms of rate and time of application is important in rose-scented geranium production. Delaying nitrogen topdressing (conventional N in the form of LAN; N 28%) after harvest to between the 7th and 9th week after cut back, was found to have a significant positive effect on biomass and essential oil production. Essential oil content of the plant did not show any response to a delay in nitrogen topdressing. A delay in nitrogen topdressing, in the first re-growth resulted into a lower citronellol to geraniol (C:G) ratio, which favour essential oil quality of the crop. Generally, the characteristics of the essential oil were within the internationally acceptable range for rose-scented geranium essential oil. In addition, production of rose-scented geranium during cooler periods is not advisable due to limited biomass production which might encourage leaching of nitrogen. The net benefits from N application is dependent on the growing period and in the present study spring and summer were more beneficial than winter and autumn. Organic N at 100 kg•ha-1•year-1 increased herbage and essential oil yield of the crop in spring and summer but further increases in organic or conventional N levels had no significant effect. N application either in winter or autumn did not improve production of the crop. Application of more N than what is required for optimum growth of the plant had no positive effect on essential oil production. Application of organic N also resulted in higher N use efficiency than conventional N. The essential oil contents (% fresh mass basis) achieved in the present study generally fell within the range of 0.04 to 0.2%. Citronellol and geraniol concentration (%), were at peak in spring season followed by summer and autumn and lowest in winter. The ratio between these two components (C:G ratio) is also used as an indicator of rose-scented geranium essential oil quality and most desirable (low C:G ratio) essential oil was attained in spring, summer and autumn harvesting seasons and least desirable oil was attained in winter. The relationship between SPAD-502 chlorophyll meter readings (SPAD units) and leaf N content (% dry weight basis) was a quadratic function. SPAD-502 chlorophyll meter readings (SPAD units) matched well with that of leaf N concentration data of rose-scented geranium. Regardless of the factors that affect the readings, this instrument can be used as an indicator of leaf N status of rose-scented geranium. Soil pH above 5.5 and soil base saturation above 55% increased fresh herbage and essential oil yield (per ha), which corresponded in this case with 2 to 6 t•ha-1 of lime application. Oil content (%) was not significantly affected by application of lime. Therefore, optimum growth of rose-scented geranium can be achieved by application of lime when plants are grown on acidic soils, but without any effect on oil content and essential oil composition. It can be concluded that N rate, source and season of production and soil pH should be considered to ensure optimal rose-scented geranium production. / Thesis (PhD)--University of Pretoria, 2012. / Plant Production and Soil Science / PhD / Unrestricted

No description available.

[pt] Esta dissertação visa compreender o fenômeno denominado pela mídia como geração canguru, através da relação entre as esferas da família e do contexto social contemporâneo, entendendo que é nesta interação que ele é construído. Esse fenômeno refere-se a uma convivência familiar prolongada na qual os filhos, jovens adultos, apesar de aptos para uma vida independente fora dos limites da casa dos pais, optam pela permanência. O adiamento da separação da família é um reflexo da conjugação de fatores intrafamiliares - ambivalência de sentimentos em relação a partida e a perda dos papéis conquistados - com extrafamiliares, fruto de um contexto social fortemente marcado por instabilidade e incerteza. Foi realizada uma pesquisa de campo com 7 famílias de classe média do Rio de Janeiro: 7 casais parentais e 8 filhos com idade a partir de 26 anos. O objetivo da pesquisa foi analisar o sentido que essas famílias imprimem a esta convivência bem como evidenciar os acordos perpetrados por seus membros em prol de prolongar o exercício de seus papéis. / [en] This thesis is an attempt to account for the phenomenon referred to by the media as the kangaroo generation in terms of the interrelations between the family sphere and the contemporary social context, on the basis of the belief that it is in this interaction that it is configured. The phenomenon in question is the fact that many young adults choose to continue living with their parents even though they already lead independent lives outside the home. The delaying of the departure from the family reflects a constellation of intrafamilial factors - ambivalence towards departure and the loss of achieved roles - and extrafamilial ones, arising from a social context sharply marked by instability and uncertainty. The field study involved seven middle-class families in Rio de Janeiro, seven parental couples and eight children aged 26 or older. The object of the research was to analyze the ways how families see this pattern of coexistence, as well as to bring out the agreements settled on by family members in order to extend in time their assigned roles.

A case study of the conservative and allegedly climate denying think tank American Enterprise Institute and its usage of climate policy delaying discourse. By applying William Lamb et al.'s model of the discourse climate delay, this essay seeks to describe and analyse if and how American Enterpise Institute uses climate delaying discourse. The study is based on published texts from the American Enterprise Institute and the analysis is performed in two steps. Firstly, by conducting an argumentation analysis using the Toulmin model on the material to discern the main arguments of the texts. Secondly, by applying Lamb et al.'s model on the arguments to discern if any of the climate policy delaying discourses are used. This study finds that almost all of Lamb et al.'s described discourses are used by the think tank and that they are used in several different ways and with different implications. This study can expand the understanding of how climate denial focused on climate policy manifests and how climate delaying discourse is used. Moreover, the results of this study further proves what previous research has indicated: That climate denialists are changing tactics and are focusing on how to affect climate policy for their own benefit.

The Tend and Befriend stress response model suggests that women have, through natural selection, evolved a different stress response reaction to that of men. It thus offers a collective, gender stereotypical reality of women’s responses to stress. In this research the Tend and Befriend model is thus viewed as a dominant public discourse which informs or influences the private narratives or stories of women. It is this interaction between public (dominant) discourses and private narratives which are investigated through using the Tend and Befriend model as a discursive landscape. If gender or gender roles are flexible, there is a concern that individual women might be misrepresented and not given a voice by the dominant discourse which supports gender stereotypical models like the Tend and Befriend model. This qualitative exploration was done by exploring the socially constructed stress responses of five professional women. To investigate this, as researcher I explored the narratives of these women in face-to-face individual interviews. The constructions explored include: How these women understand the way they respond to stress; how they view the Tend and Befriend model; and the influence of the model on them. Through the lenses of social constructionism a broader insight into the stress responses of women may be obtained. From the data analysis, I uncovered very little ‘evidence’ for tending or befriending behaviour as described by Taylor, Klein, Lewis, Gruenewald, Gurung and Updegraff (2000), with the participants. In the exploration the closest response to the model which the participants reported was befriending, however in their construction of befriending they employed it as a workplace strategy. The only form of tending co-constructed in the interview process was a secondary response to stress and a unique outcome to this study: Self-tending. Additionally, as social constructionist research predicts, these participants illustrated that for them stress responses are not concrete, as models would like to suggest, rather they employed an alternate multifaceted stress response approach which was another significant unique outcome to this study. / Dissertation (MA)--University of Pretoria, 2010. / Psychology / unrestricted

A Internet tem se tornado uma plataforma importante para interação e compartilhamento de arquivos, o que motivou uma crescente demanda por serviços eficientes. Sistemas de distribuição de conteúdo (CDS) precisaram ser criados visando modernidade e robustez. No contexto desta tese, CDS são definidos como sistemas usados para compartilhar qualquer tipo de conteúdo na Internet. Duas categorias de CDS se destacam como as mais populares: compartilhamento de arquivos e sistemas de mídia contínua. Arquiteturas par-a-par (P2P) surgiram como potenciais soluções para o aprimoramento da disseminação de conteúdo nos CDS. Nesse contexto, a popularização das arquiteturas P2P motivou a comunidade científica a investigar alguns aspectos de pesquisa desafiadores, e.g., otimização de topologias de redes, mecanismos de inicialização de sistemas e serviços de descoberta de recursos. Um desafio com interesse especial a esta tese diz respeito a mecanismos para conciliar a preferência dos usuários aos conteúdos publicados. Esse aspecto é importante para garantir uma boa qualidade de experiência (QoE) aos usuários dos sistemas, uma vez que podem existir divergências entre opiniões na descrição dos conteúdos e ações maliciosas. Esforços de pesquisa constantes têm sido feitos para combater poluição de conteúdo em CDS. Abordagens buscam construir uma base de conhecimento sobre poluidores e conteúdos poluídos para identificar e isolar conteúdos suspeitos depois que eles são publicados. Entretanto, o tempo de reação dessas abordagens até considerar um conteúdo poluído é consideravelmente longo, permitindo uma ampla disseminação de poluição. Além disso, algumas abordagens anteriores buscam polarizar conteúdos entre poluídos ou não, desconsiderando a intrínseca subjetividade acerca da classificação dos conteúdos compartilhados. O objetivo principal desta tese é propor um mecanismo para prover uma boa QoE aos usuários – agindo proativamente durante as fases iniciais da publicação dos conteúdos – e reduzir os efeitos de interferências maliciosas. Para alcançar tal objetivo, três passos principais guiaram o trabalho de pesquisa apresentado nesta tese. Primeiro, propusemos uma estratégia inovadora que opera de forma conservadora para conter a disseminação de poluição. Segundo, estendemos nossa solução para lidar com a subjetividade acerca das descrições dos conteúdos. Terceiro, tratamos o ataque de poluição como um ataque massivo. Para avaliar a solução, experimentos foram executados utilizando testes reais e simulações. Resultados ressaltaram a importância de adotar medidas de segurança para combater comportamentosmaliciosos em CDS. Na ausência de mecanismos de contramedida, pequenas proporções (10%) de atacantes foram capazes de comprometer o sistema. A instanciação da estratégia conservadora proposta nesta tese demonstrou a eficácia em atrasar usuários para contornar ataques massivos. / The Internet has become a large platform where users can interact and share personal files or third-party productions. Considering the increasing demand for efficient content sharing, modern and robust content distribution systems (CDS) need to be deployed and maintained. In the context of this thesis, CDS are defined as systems used for sharing any kind of content on the Internet. Two categories of CDS are underscored as the most popular ones: file sharing and streaming systems. Peer-to-peer (P2P) architectures have emerged as a potential solution to improve content dissemination in CDS. The popularization of P2P architectures, in the context of CDS,motivated the scientific community to investigate some challenging problems, namely network topology optimization, bootstrap mechanisms, and service discovery. One particular interesting challenge, in the context of this thesis, is related to mechanisms to approximate users to their personal interests. This is important to guarantee good quality of experience (QoE) to users when searching for content. Imprecise descriptions are likely to happen due to different users’ opinion or malicious behavior. Substantial research has been carried out to fight content pollution in CDS. Proposed approaches try to identify and isolate suspicious content after publication. The rationale is to build a base of knowledge about polluters and fake content. However, the reaction time until a content is considered polluted is considerably long, which allows pollution to get widely disseminated. Furthermore, some previous approaches attempt to polarize contents in either polluted or not, not taking into account the inherent subjectivity behind the evaluation of shared contents. The main objective of this thesis is to devise a mechanism to provide users a good QoE – by acting proactively in the early stages of content distribution life cycle – and reduce the effect of malicious interferences. To achieve that, three main steps guided the research work presented in this thesis. First, we proposed a novel strategy that operates conservatively to avoid wide pollution dissemination. Second, we extended our previous solution to cope with the subjectivity regarding content descriptions. Third, and last, we address the pollution attack as a massive attack. To evaluate our solution, a set of experiments was carried out using both real tests and simulations. Results showed the importance of adopting security measures to mitigate malicious behavior in CDS. In the absence of countermeasure mechanisms, even a small proportion (10%) of attackers was able to subvert the system. The introduction of a conservative strategy in this thesis demonstrated the efficacy of delaying users in circumventing massive attacks.

A Internet tem se tornado uma plataforma importante para interação e compartilhamento de arquivos, o que motivou uma crescente demanda por serviços eficientes. Sistemas de distribuição de conteúdo (CDS) precisaram ser criados visando modernidade e robustez. No contexto desta tese, CDS são definidos como sistemas usados para compartilhar qualquer tipo de conteúdo na Internet. Duas categorias de CDS se destacam como as mais populares: compartilhamento de arquivos e sistemas de mídia contínua. Arquiteturas par-a-par (P2P) surgiram como potenciais soluções para o aprimoramento da disseminação de conteúdo nos CDS. Nesse contexto, a popularização das arquiteturas P2P motivou a comunidade científica a investigar alguns aspectos de pesquisa desafiadores, e.g., otimização de topologias de redes, mecanismos de inicialização de sistemas e serviços de descoberta de recursos. Um desafio com interesse especial a esta tese diz respeito a mecanismos para conciliar a preferência dos usuários aos conteúdos publicados. Esse aspecto é importante para garantir uma boa qualidade de experiência (QoE) aos usuários dos sistemas, uma vez que podem existir divergências entre opiniões na descrição dos conteúdos e ações maliciosas. Esforços de pesquisa constantes têm sido feitos para combater poluição de conteúdo em CDS. Abordagens buscam construir uma base de conhecimento sobre poluidores e conteúdos poluídos para identificar e isolar conteúdos suspeitos depois que eles são publicados. Entretanto, o tempo de reação dessas abordagens até considerar um conteúdo poluído é consideravelmente longo, permitindo uma ampla disseminação de poluição. Além disso, algumas abordagens anteriores buscam polarizar conteúdos entre poluídos ou não, desconsiderando a intrínseca subjetividade acerca da classificação dos conteúdos compartilhados. O objetivo principal desta tese é propor um mecanismo para prover uma boa QoE aos usuários – agindo proativamente durante as fases iniciais da publicação dos conteúdos – e reduzir os efeitos de interferências maliciosas. Para alcançar tal objetivo, três passos principais guiaram o trabalho de pesquisa apresentado nesta tese. Primeiro, propusemos uma estratégia inovadora que opera de forma conservadora para conter a disseminação de poluição. Segundo, estendemos nossa solução para lidar com a subjetividade acerca das descrições dos conteúdos. Terceiro, tratamos o ataque de poluição como um ataque massivo. Para avaliar a solução, experimentos foram executados utilizando testes reais e simulações. Resultados ressaltaram a importância de adotar medidas de segurança para combater comportamentosmaliciosos em CDS. Na ausência de mecanismos de contramedida, pequenas proporções (10%) de atacantes foram capazes de comprometer o sistema. A instanciação da estratégia conservadora proposta nesta tese demonstrou a eficácia em atrasar usuários para contornar ataques massivos. / The Internet has become a large platform where users can interact and share personal files or third-party productions. Considering the increasing demand for efficient content sharing, modern and robust content distribution systems (CDS) need to be deployed and maintained. In the context of this thesis, CDS are defined as systems used for sharing any kind of content on the Internet. Two categories of CDS are underscored as the most popular ones: file sharing and streaming systems. Peer-to-peer (P2P) architectures have emerged as a potential solution to improve content dissemination in CDS. The popularization of P2P architectures, in the context of CDS,motivated the scientific community to investigate some challenging problems, namely network topology optimization, bootstrap mechanisms, and service discovery. One particular interesting challenge, in the context of this thesis, is related to mechanisms to approximate users to their personal interests. This is important to guarantee good quality of experience (QoE) to users when searching for content. Imprecise descriptions are likely to happen due to different users’ opinion or malicious behavior. Substantial research has been carried out to fight content pollution in CDS. Proposed approaches try to identify and isolate suspicious content after publication. The rationale is to build a base of knowledge about polluters and fake content. However, the reaction time until a content is considered polluted is considerably long, which allows pollution to get widely disseminated. Furthermore, some previous approaches attempt to polarize contents in either polluted or not, not taking into account the inherent subjectivity behind the evaluation of shared contents. The main objective of this thesis is to devise a mechanism to provide users a good QoE – by acting proactively in the early stages of content distribution life cycle – and reduce the effect of malicious interferences. To achieve that, three main steps guided the research work presented in this thesis. First, we proposed a novel strategy that operates conservatively to avoid wide pollution dissemination. Second, we extended our previous solution to cope with the subjectivity regarding content descriptions. Third, and last, we address the pollution attack as a massive attack. To evaluate our solution, a set of experiments was carried out using both real tests and simulations. Results showed the importance of adopting security measures to mitigate malicious behavior in CDS. In the absence of countermeasure mechanisms, even a small proportion (10%) of attackers was able to subvert the system. The introduction of a conservative strategy in this thesis demonstrated the efficacy of delaying users in circumventing massive attacks.

A Internet tem se tornado uma plataforma importante para interação e compartilhamento de arquivos, o que motivou uma crescente demanda por serviços eficientes. Sistemas de distribuição de conteúdo (CDS) precisaram ser criados visando modernidade e robustez. No contexto desta tese, CDS são definidos como sistemas usados para compartilhar qualquer tipo de conteúdo na Internet. Duas categorias de CDS se destacam como as mais populares: compartilhamento de arquivos e sistemas de mídia contínua. Arquiteturas par-a-par (P2P) surgiram como potenciais soluções para o aprimoramento da disseminação de conteúdo nos CDS. Nesse contexto, a popularização das arquiteturas P2P motivou a comunidade científica a investigar alguns aspectos de pesquisa desafiadores, e.g., otimização de topologias de redes, mecanismos de inicialização de sistemas e serviços de descoberta de recursos. Um desafio com interesse especial a esta tese diz respeito a mecanismos para conciliar a preferência dos usuários aos conteúdos publicados. Esse aspecto é importante para garantir uma boa qualidade de experiência (QoE) aos usuários dos sistemas, uma vez que podem existir divergências entre opiniões na descrição dos conteúdos e ações maliciosas. Esforços de pesquisa constantes têm sido feitos para combater poluição de conteúdo em CDS. Abordagens buscam construir uma base de conhecimento sobre poluidores e conteúdos poluídos para identificar e isolar conteúdos suspeitos depois que eles são publicados. Entretanto, o tempo de reação dessas abordagens até considerar um conteúdo poluído é consideravelmente longo, permitindo uma ampla disseminação de poluição. Além disso, algumas abordagens anteriores buscam polarizar conteúdos entre poluídos ou não, desconsiderando a intrínseca subjetividade acerca da classificação dos conteúdos compartilhados. O objetivo principal desta tese é propor um mecanismo para prover uma boa QoE aos usuários – agindo proativamente durante as fases iniciais da publicação dos conteúdos – e reduzir os efeitos de interferências maliciosas. Para alcançar tal objetivo, três passos principais guiaram o trabalho de pesquisa apresentado nesta tese. Primeiro, propusemos uma estratégia inovadora que opera de forma conservadora para conter a disseminação de poluição. Segundo, estendemos nossa solução para lidar com a subjetividade acerca das descrições dos conteúdos. Terceiro, tratamos o ataque de poluição como um ataque massivo. Para avaliar a solução, experimentos foram executados utilizando testes reais e simulações. Resultados ressaltaram a importância de adotar medidas de segurança para combater comportamentosmaliciosos em CDS. Na ausência de mecanismos de contramedida, pequenas proporções (10%) de atacantes foram capazes de comprometer o sistema. A instanciação da estratégia conservadora proposta nesta tese demonstrou a eficácia em atrasar usuários para contornar ataques massivos. / The Internet has become a large platform where users can interact and share personal files or third-party productions. Considering the increasing demand for efficient content sharing, modern and robust content distribution systems (CDS) need to be deployed and maintained. In the context of this thesis, CDS are defined as systems used for sharing any kind of content on the Internet. Two categories of CDS are underscored as the most popular ones: file sharing and streaming systems. Peer-to-peer (P2P) architectures have emerged as a potential solution to improve content dissemination in CDS. The popularization of P2P architectures, in the context of CDS,motivated the scientific community to investigate some challenging problems, namely network topology optimization, bootstrap mechanisms, and service discovery. One particular interesting challenge, in the context of this thesis, is related to mechanisms to approximate users to their personal interests. This is important to guarantee good quality of experience (QoE) to users when searching for content. Imprecise descriptions are likely to happen due to different users’ opinion or malicious behavior. Substantial research has been carried out to fight content pollution in CDS. Proposed approaches try to identify and isolate suspicious content after publication. The rationale is to build a base of knowledge about polluters and fake content. However, the reaction time until a content is considered polluted is considerably long, which allows pollution to get widely disseminated. Furthermore, some previous approaches attempt to polarize contents in either polluted or not, not taking into account the inherent subjectivity behind the evaluation of shared contents. The main objective of this thesis is to devise a mechanism to provide users a good QoE – by acting proactively in the early stages of content distribution life cycle – and reduce the effect of malicious interferences. To achieve that, three main steps guided the research work presented in this thesis. First, we proposed a novel strategy that operates conservatively to avoid wide pollution dissemination. Second, we extended our previous solution to cope with the subjectivity regarding content descriptions. Third, and last, we address the pollution attack as a massive attack. To evaluate our solution, a set of experiments was carried out using both real tests and simulations. Results showed the importance of adopting security measures to mitigate malicious behavior in CDS. In the absence of countermeasure mechanisms, even a small proportion (10%) of attackers was able to subvert the system. The introduction of a conservative strategy in this thesis demonstrated the efficacy of delaying users in circumventing massive attacks.

Dans de nombreux domaines dans lesquels les mesures ou les données sont disponibles séquentiellement, il est important de savoir décider le plus tôt possible, même si c’est à partir d’informations encore incomplètes. C’est le cas par exemple en milieu hospitalier où l’apprentissage de règles de décision peut se faire à partir de cas complètement documentés, mais où, devant un nouveau patient, il peut être crucial de prendre une dé- cision très rapidement. Dans ce type de contextes, un compromis doit être optimisé entre la possibilité d’arriver à une meilleure décision en attendant des mesures supplé- mentaires, et le coût croissant associé à chaque nouvelle mesure. Nous considérons dans cette thèse un nouveau cadre général de classification précoce de séries temporelles où le coût d’attente avant de prendre une décision est explicitement pris en compte lors de l’optimisation du compromis entre la qualité et la précocité de prédictions. Nous proposons donc un critère formel qui exprime ce compromis, ainsi que deux approches différentes pour le résoudre. Ces approches sont intéressantes et apportent deux propriétés désirables pour décider en ligne : (i) elles estiment en ligne l’instant optimal dans le futur où une minimisation du critère peut être prévue. Elles vont donc au-delà des approches classiques qui décident d’une façon myope, à chaque instant, d’émettre une prédiction ou d’attendre plus d’information, (ii) ces approches sont adaptatives car elles prennent en compte les propriétés de la série temporelle en entrée pour estimer l’instant optimal pour la classifier. Des expériences extensives sur des données contrôlées et sur des données réelles montrent l’intérêt de ces approches pour fournir des prédictions précoces, fiables, adaptatives et non myopes, ce qui est indispensable dans de nombreuses applications. / Early classification of time series is becoming increasingly a valuable task for assisting in decision making process in many application domains. In this setting, information can be gained by waiting for more evidences to arrive, thus helping to make better decisions that incur lower misclassification costs, but, meanwhile, the cost associated with delaying the decision generally increases, rendering the decision less attractive. Making early predictions provided that are accurate requires then to solve an optimization problem combining two types of competing costs. This thesis introduces a new general framework for time series early classification problem. Unlike classical approaches that implicitly assume that misclassification errors are cost equally and the cost of delaying the decision is constant over time, we cast the the problem as a costsensitive online decision making problem when delaying the decision is costly. We then propose a new formal criterion, along with two approaches that estimate the optimal decision time for a new incoming yet incomplete time series. In particular, they capture the evolutions of typical complete time series in the training set thanks to a segmentation technique that forms meaningful groups, and leverage these complete information to estimate the costs for all future time steps where data points still missing. These approaches are interesting in two ways: (i) they estimate, online, the earliest time in the future where a minimization of the criterion can be expected. They thus go beyond the classical approaches that myopically decide at each time step whether to make a decision or to postpone the call one more time step, and (ii) they are adaptive, in that the properties of the incoming time series are taken into account to decide when is the optimal time to output a prediction. Results of extensive experiments on synthetic and real data sets show that both approaches successfully meet the behaviors expected from early classification systems.