• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 51
  • 31
  • 19
  • 18
  • 12
  • 3
  • 3
  • 2
  • 2
  • 1
  • Tagged with
  • 160
  • 33
  • 18
  • 18
  • 16
  • 16
  • 16
  • 14
  • 13
  • 12
  • 11
  • 11
  • 11
  • 10
  • 10
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
71

From confusion noise to active learning : playing on label availability in linear classification problems / Du bruit de confusion à l’apprentissage actif : jouer sur la disponibilité des étiquettes dans les problèmes de classification linéaire

Louche, Ugo 04 July 2016 (has links)
Les travaux présentés dans cette thèse relèvent de l'étude des méthodes de classification linéaires, c'est à dire l'étude de méthodes ayant pour but la catégorisation de données en différents groupes à partir d'un jeu d'exemples, préalablement étiquetés, disponible en amont et appelés ensemble d'apprentissage. En pratique, l'acquisition d'un tel ensemble d'apprentissage peut être difficile et/ou couteux, la catégorisation d'un exemple étant de fait plus ardu que l'obtention de dudit exemple. Cette disparité entre la disponibilité des données et notre capacité à constituer un ensemble d'apprentissage étiqueté a été un des problèmes centraux de l'apprentissage automatique et ce manuscrit s’intéresse à deux solutions usuellement considérées pour contourner ce problème : l'apprentissage en présence de données bruitées et l'apprentissage actif. / The works presented in this thesis fall within the general framework of linear classification, that is the problem of categorizing data into two or more classes based on on a training set of labelled data. In practice though acquiring labeled examples might prove challenging and/or costly as data are inherently easier to obtain than to label. Dealing with label scarceness have been a motivational goal in the machine learning literature and this work discuss two settings related to this problem: learning in the presence of noise and active learning.
72

Méthodes ensembliste pour des problèmes de classification multi-vues et multi-classes avec déséquilibres / Tackling the uneven views problem with cooperation based ensemble learning methods

Koco, Sokol 16 December 2013 (has links)
De nos jours, dans plusieurs domaines, tels que la bio-informatique ou le multimédia, les données peuvent être représentées par plusieurs ensembles d'attributs, appelés des vues. Pour une tâche de classification donnée, nous distinguons deux types de vues : les vues fortes sont celles adaptées à la tâche, les vues faibles sont adaptées à une (petite) partie de la tâche ; en classification multi-classes, chaque vue peut s'avérer forte pour reconnaître une classe, et faible pour reconnaître d’autres classes : une telle vue est dite déséquilibrée. Les travaux présentés dans cette thèse s'inscrivent dans le cadre de l'apprentissage supervisé et ont pour but de traiter les questions d'apprentissage multi-vue dans le cas des vues fortes, faibles et déséquilibrées. La première contribution de cette thèse est un algorithme d'apprentissage multi-vues théoriquement fondé sur le cadre de boosting multi-classes utilisé par AdaBoost.MM. La seconde partie de cette thèse concerne la mise en place d'un cadre général pour les méthodes d'apprentissage de classes déséquilibrées (certaines classes sont plus représentées que les autres). Dans la troisième partie, nous traitons le problème des vues déséquilibrées en combinant notre approche des classes déséquilibrées et la coopération entre les vues mise en place pour appréhender la classification multi-vues. Afin de tester les méthodes sur des données réelles, nous nous intéressons au problème de classification d'appels téléphoniques, qui a fait l'objet du projet ANR DECODA. Ainsi chaque partie traite différentes facettes du problème. / Nowadays, in many fields, such as bioinformatics or multimedia, data may be described using different sets of features, also called views. For a given classification task, we distinguish two types of views:strong views, which are suited for the task, and weak views suited for a (small) part of the task; in multi-class learning, a view can be strong with respect to some (few) classes and weak for the rest of the classes: these are imbalanced views. The works presented in this thesis fall in the supervised learning setting and their aim is to address the problem of multi-view learning under strong, weak and imbalanced views, regrouped under the notion of uneven views. The first contribution of this thesis is a multi-view learning algorithm based on the same framework as AdaBoost.MM. The second part of this thesis proposes a unifying framework for imbalanced classes supervised methods (some of the classes are more represented than others). In the third part of this thesis, we tackle the uneven views problem through the combination of the imbalanced classes framework and the between-views cooperation used to take advantage of the multiple views. In order to test the proposed methods on real-world data, we consider the task of phone calls classifications, which constitutes the subject of the ANR DECODA project. Each part of this thesis deals with different aspects of the problem.
73

Reality check : inferential confusion and cognitive confidence as core cognitive factors across the obsessive-compulsive spectrum

Ouellet-Courtois, Catherine 08 1900 (has links)
Le trouble obsessionnel-compulsif (TOC) se caractérise par la présence d’obsessions et/ou de compulsions. À la lumière de l’hétérogénéité du TOC et de la présence de styles de pensées et de comportements de type TOC chez des personnes présentant d’autres problèmes de santé mentale, certains ont fait valoir la nécessité de créer une catégorie des troubles du spectre obsessionnel-compulsif et d’identifier les processus cognitifs communs qui sous-tendent ces troubles afin d’élaborer des théories et des traitements plus parcimonieux. Une tendance générale à douter de ses sens et de ses facultés cognitives semble être le pivot des troubles obsessionnels. Selon l’approche basée sur les inférences, le doute obsessionnel est suscité par un processus de raisonnement erroné, soit la confusion inférentielle (CI). La CI implique (1) une méfiance vis-à-vis des sens et (2) une importance indue accordée aux possibilités imaginaires. La faible confiance cognitive (CC), un processus cognitif similaire, renvoie à une méfiance par rapport à son attention, sa perception et sa mémoire. Cette thèse a visé à étudier la CI et la faible CC en tant que potentiels facteurs cognitifs transdiagnostiques dans le spectre de l’obsessionnalité. Le premier article constitue une revue systématique avec méta-analyse destinée à évaluer le rôle de la CC pour différents sous-types du TOC et à examiner à quel degré la faible CC est associée aux symptômes du TOC. On a constaté que les individus atteints d’un TOC présentent une plus faible CC que les témoins sains, mais que celle-ci ne semble pas spécifique au TOC. L’article a aussi souligné la nécessité d’employer des tâches idiosyncratiques, ciblant les distorsions de la pensée propres au TOC, afin de bien mesurer la CC. Dans le cadre du deuxième article, le but a été d’approfondir cette piste de recherche en examinant le rôle commun de la faible CC et de la CI pour les différents sous-types du TOC, en procédant à des analyses de grappes avec un échantillon de 128 patients atteints d’un TOC. Alors qu’il a été constaté que la faible CC correspondait davantage aux sous-types de vérification et « tout à fait juste », la CI semble pertinente pour un plus large éventail de profils TOC. Le troisième article examine le rôle de la CI chez les troubles des conduites alimentaires (TCAs) en provoquant la CI expérimentalement. Des participantes atteintes d’un TA (n = 18) et des femmes témoins saines (n = 18) ont été assignées à l’une des deux conditions expérimentales : pour la condition CI élevée, les participantes ont visionné des vidéos où des séquences clés étaient manquantes, ce qui suscitait la CI; pour la condition CI faible, les participantes ont visionné les vidéos intégrales. Chez le groupe TA assigné à la condition CI élevée, on a observé une tendance à présenter un état de CI post-vidéos supérieur, un recours accru au comportement de neutralisation et, enfin, davantage de symptômes TOC. En somme, les résultats de cette thèse soulignent la pertinence de la CI et de la faible CC en tant que facteurs cognitifs transdiagnostiques sur le spectre obsessionnel-compulsif. / Obsessive-compulsive disorder (OCD) is a severe mental health disorder that involves obsessions and/or compulsions. In light of the heterogeneity of OCD and of the presence of OCD-like thinking and behaviors in several disorders, some have argued for the necessity of a new category of obsessive-compulsive spectrum disorders. Considering the overlap between various disorders characterized by obsessionality, there is a need for the identification of common cognitive processes that underpin these disorders in order to formulate more parsimonious explanations and treatments for these conditions. A general tendency to doubt the senses and cognitive faculties appears as central to obsessional disorders. According to the inference-based approach, the obsessional doubt is elicited by a faulty reasoning process known as inferential confusion (IC), that implicates (1) a distrust of the senses, and (2) an investment in imaginary possibilities. A similar construct is low cognitive confidence, which is defined as a distrust of one’s attention, perception and memory. The overarching goal of this thesis was to examine IC and low cognitive confidence as potential transdiagnostic cognitive factors across the spectrum of obsessionality. The first thesis article evaluated the role of cognitive confidence across OCD subtypes and examined the extent to which poor cognitive confidence is associated with OCD symptomatology by conducting a systematic review with a meta-analysis. This article led to the conclusion that individuals with OCD have lower cognitive confidence than healthy controls, but that it is unclear if cognitive confidence is specific to OCD, such that the use of idiosyncratic tasks appears to be required in order to correctly capture cognitive confidence in OCD. The second thesis article furthered this investigation by examining the joint role of low cognitive confidence and IC across OCD subtypes by performing cluster analyses using a sample of 128 individuals with OCD. While cognitive confidence was found to more relevant to the checking and “just right” subtypes, IC appeared to be relevant to a wider range of OCD profiles. The third thesis article sought to examine the role of IC in eating disorders by inducing IC experimentally. Female participants with an eating disorder (n = 18) and healthy controls participants (n = 18) were recruited. Participants were assigned to one of two experimental conditions: in the High IC condition, participants watched videos with key sequences missing – provoking a distrust of the senses and lending more space for the imagination, thus triggering IC. In the Low IC condition, participants watched videos without sequences missing. The eating disorder group assigned to the High IC condition demonstrated a trend for higher levels of state IC, greater neutralization behavior and higher OCD symptoms than those who were assigned to the Low IC condition. Taken together, the results of the present thesis underline the relevance of IC and cognitive confidence as transdiagnostic cognitive factors across the obsessive-compulsive spectrum.
74

Factors affecting the performance of trainable models for software defect prediction

Bowes, David Hutchinson January 2013 (has links)
Context. Reports suggest that defects in code cost the US in excess of $50billion per year to put right. Defect Prediction is an important part of Software Engineering. It allows developers to prioritise the code that needs to be inspected when trying to reduce the number of defects in code. A small change in the number of defects found will have a significant impact on the cost of producing software. Aims. The aim of this dissertation is to investigate the factors which a ect the performance of defect prediction models. Identifying the causes of variation in the way that variables are computed should help to improve the precision of defect prediction models and hence improve the cost e ectiveness of defect prediction. Methods. This dissertation is by published work. The first three papers examine variation in the independent variables (code metrics) and the dependent variable (number/location of defects). The fourth and fifth papers investigate the e ect that di erent learners and datasets have on the predictive performance of defect prediction models. The final paper investigates the reported use of di erent machine learning approaches in studies published between 2000 and 2010. Results. The first and second papers show that independent variables are sensitive to the measurement protocol used, this suggests that the way data is collected a ects the performance of defect prediction. The third paper shows that dependent variable data may be untrustworthy as there is no reliable method for labelling a unit of code as defective or not. The fourth and fifth papers show that the dataset and learner used when producing defect prediction models have an e ect on the performance of the models. The final paper shows that the approaches used by researchers to build defect prediction models is variable, with good practices being ignored in many papers. Conclusions. The measurement protocols for independent and dependent variables used for defect prediction need to be clearly described so that results can be compared like with like. It is possible that the predictive results of one research group have a higher performance value than another research group because of the way that they calculated the metrics rather than the method of building the model used to predict the defect prone modules. The machine learning approaches used by researchers need to be clearly reported in order to be able to improve the quality of defect prediction studies and allow a larger corpus of reliable results to be gathered.
75

Análise geoestatística de mapas temáticos da produtividade da soja com diferentes grades amostrais / Geostatistical analysis of thematic maps of soybean yield with differente sampling grids

Kestring, Franciele Buss Frescki 07 July 2011 (has links)
Made available in DSpace on 2017-07-10T19:24:53Z (GMT). No. of bitstreams: 1 Franciele_texto.pdf: 972546 bytes, checksum: 4159555de632249d0c83764a3aecc74c (MD5) Previous issue date: 2011-07-07 / Studies on spatial variability of soybeans yield are of great importance for the development of new technologies that improve the world agricultural production. One of methods that allows this study is geostatistics. The geostatistical analysis makes possible the predictions of results and one of its products are thematic maps. Thus, this trial describes some techniques to draw and compare thematic maps using kriging. The analysis was based on data from soybean yield in t ha−1 according to harvest year 2004/2005 in an experimental area with sampling grades whose distances were: 25x25 m, 50x50 m, 75x75 m and 100x100 m plus a harvest monitor. The maps were compared using error matrix and confusion matrix. In addition, there was a better accuracy of the spatial variability maps that were drawn, while the analysis of coefficients of accuracy allows a better planning of sampling mesh for future studies. The measures of accuracy that were obtained by error matrix are significant options to make comparison among thematic maps, once they provide global indices and also by classes. / Com o aumento da produção agrícola mundial, o processo de produção agrícola tornou-se alvo do estudo de diversos pesquisadores. Estudos sobre a variabilidade espacial da produtividade da soja são de grande importância para o desenvolvimento de novas tecnologias, que beneficiam a agricultura. A análise geoestatística torna possível realizar previsões dos resultados, tendo como um de seus produtos os mapas temáticos. Este trabalho descreve algumas técnicas para a construção e comparação de mapas temáticos, utilizando a krigagem. A análise foi realizada com dados da produtividade de soja em t ha−1 do ano agrícola 2004/2005 numa área experimental com grades de amostragem com distâncias de 25x25 m, 50x50 m, 75x75 m, 100x100 m e monitor de colheita, comparando-se os mapas, utilizando a matriz de erros e a matriz de confusão. Além de uma melhor precisão dos mapas de variabilidade espacial gerados, a análise dos índices de acurácia possibilita um melhor planejamento das malhas amostrais para futuros estudos. As medidas de acurácia obtidas por meio da matriz de erros são opções significativas para realizar a comparação entre mapas temáticos, uma vez que fornecem índices globais e também por classes.
76

Análise geoestatística de mapas temáticos da produtividade da soja com diferentes grades amostrais / Geostatistical analysis of thematic maps of soybean yield with differente sampling grids

Kestring, Franciele Buss Frescki 07 July 2011 (has links)
Made available in DSpace on 2017-05-12T14:48:16Z (GMT). No. of bitstreams: 1 Franciele_texto.pdf: 972546 bytes, checksum: 4159555de632249d0c83764a3aecc74c (MD5) Previous issue date: 2011-07-07 / Studies on spatial variability of soybeans yield are of great importance for the development of new technologies that improve the world agricultural production. One of methods that allows this study is geostatistics. The geostatistical analysis makes possible the predictions of results and one of its products are thematic maps. Thus, this trial describes some techniques to draw and compare thematic maps using kriging. The analysis was based on data from soybean yield in t ha−1 according to harvest year 2004/2005 in an experimental area with sampling grades whose distances were: 25x25 m, 50x50 m, 75x75 m and 100x100 m plus a harvest monitor. The maps were compared using error matrix and confusion matrix. In addition, there was a better accuracy of the spatial variability maps that were drawn, while the analysis of coefficients of accuracy allows a better planning of sampling mesh for future studies. The measures of accuracy that were obtained by error matrix are significant options to make comparison among thematic maps, once they provide global indices and also by classes. / Com o aumento da produção agrícola mundial, o processo de produção agrícola tornou-se alvo do estudo de diversos pesquisadores. Estudos sobre a variabilidade espacial da produtividade da soja são de grande importância para o desenvolvimento de novas tecnologias, que beneficiam a agricultura. A análise geoestatística torna possível realizar previsões dos resultados, tendo como um de seus produtos os mapas temáticos. Este trabalho descreve algumas técnicas para a construção e comparação de mapas temáticos, utilizando a krigagem. A análise foi realizada com dados da produtividade de soja em t ha−1 do ano agrícola 2004/2005 numa área experimental com grades de amostragem com distâncias de 25x25 m, 50x50 m, 75x75 m, 100x100 m e monitor de colheita, comparando-se os mapas, utilizando a matriz de erros e a matriz de confusão. Além de uma melhor precisão dos mapas de variabilidade espacial gerados, a análise dos índices de acurácia possibilita um melhor planejamento das malhas amostrais para futuros estudos. As medidas de acurácia obtidas por meio da matriz de erros são opções significativas para realizar a comparação entre mapas temáticos, uma vez que fornecem índices globais e também por classes.
77

Relecture des multiples facettes du féminin sacré et profane

Chauvin, Marilyn 27 June 2012 (has links) (PDF)
L'Art a toujours consacré une grande part à l'image du féminin. Que cela soit dans l'iconographie gréco-romaine ou judéo-chrétienne, ses multiples représentations sont synonymes de confusion et d'ambivalence. La femme et l'image ont en commun de susciter méfiance et fascination. C'est au travers de l'étude approfondie de quelques figures clefs de l'histoire de l'art, que nous vous proposons une relecture post-féministe des diverses facettes du féminin, prises dans la dyade sacré/profane. Mythes et croyances donnèrent naissance à un métissage pagano-chrétien qui fit émerger un Eternel Féminin inébranlable encore très prégnant dans l'art actuel. Tantôt dans la foi de l'image, tantôt dans sa condamnation, ainsi se résume l'insoluble combat entre l'humain et le divin. La Femme restera à jamais l'élément trouble associé au paraître et à la beauté. Dans nos recherches nous avons constaté que la femme et la peinture sont en parfaite adéquation. Elles sont indissociables car iconoclasme et misogynie vont souvent de pair. Peu à peu, le corps remplacera la toile et le fard, la peinture pour les artistes transgenres. Dans d'étranges (queer) parodies entre exhibition et chamanisme, ils réinventeront leur devenir-féminin. La surface de l'œuvre devient alors le miroir où se reflète cet Autre, alter ego tant recherché. La pensée féministe se met en marche au travers des révolutions des genres et des sexes. Ainsi, c'est entre Pygmalion et Narcisse que les artistes des XXéme et XXIéme siècles, nous offrent, entre mascarade et mélancolie, la vision d'un idéal sans original. Enfin, nous revenons sur l'art des femmes. Leurs pratiques sont souvent borderline, partagées entre violence, humour et Charis dans leur quête d'un sacré hors religion. Elles offrent un devenir-pandorien de l'art pour briser à jamais le désormais trop célèbre : " Sois belle et tais toi! "
78

Två sidor av samma mynt : Brand alliance bidrar till brand confusion i bankbranschen

Andhäll, David, Thörn, Patrik January 2013 (has links)
Bergslagens Sparbank as well as other independent(fristående) sparbanker have discoveredthat some of their customers can not separate them from Swedbank. Some customers maythus believe that they are the same bank. This study seeks factors that may contribute to thisproblem, and measures that can counteract. We refer to this problem as brand confusion andmainly have our focus on factors derived from brand alliance, but also from the companyitself and the banking industry. We use a qualitative method where we collect empirical datafrom interviews with four sparbanker. Strategic differences between them, allow us to seehow different factors can contribute to and or counteract brand confusion. This study showsthat the graphic profile, that includes the common symbol and colors, may have great impactwhen it comes to sparbankerna. However, the messy business history that includes thecomplex partner relationship could also have a great impact. Finally, we would also like tomention the importance of how all involving parties in a brand alliance communicates this. Ifit does not come through loud and clear to the customer that there is a partnership, strongsimilarities and ambiguities may affect customers to believe two companies are one and underthe same brand.
79

The Confusion Doctrine; Establishing Swedish compliance with EU Law

Eriksson, Rebecca January 2010 (has links)
As a response to trade marks’ enhanced importance within trade, the EU’s interest in the area has increased by proponing a harmonization of the member states’ trade mark pro-tection so far as needed to preserve the EU’s objective of an internal market. The area is therefore regulated by an EU Directive, however allowing some national discretion. The purpose of this study was to investigate if a specific part of the trade mark protec-tion, the assessment-based confusion doctrine, corresponds on a Swedish and EU level. The aim was to locate statutory discrepancies in order to stimulate further review of the practical application of the doctrine from the analytical perspective of legal certainty. A scientifically accepted and traditional legal research method was applied when ex-amining and interpreting the sources of law. In addition, a comparative study was con-ducted between the two investigated legal systems to achieve the overall purpose. When comparing the results from the investigated sources, the legislations present a sta-tutory diversity, opening up for practical discrepancies. So was also the case with the application at the early stage of national implementation of the EU Directive. The tradi-tional national confusion doctrine, prescribing a more legal-technical assessment, did not correspond to the more flexible and contemporary EU view. Consequently, some national courts had to endure criticism for not adjusting to the EU development. Later case law however presents a very positive transition to the EU view of the confu-sion doctrine, suggesting a partial abandonment of the national legal sources of law for the benefit of EU law. Conclusion was however that despite this practical transition to EU law, statutory changes are necessary in order to safeguard the legal certainty in the way of achieving predictability.
80

How reliable are the marginal totals in cooperation experiments in the laboratory?

Berger, Roger 22 July 2014 (has links) (PDF)
Subjects in laboratory experiments are prone to effects of social desirability. This reactive behavior is due to the subjects perceived anonymity in the entire experiment. Especially, socially desirable behavior is also triggered by assembling and/or payment procedures that are not anonymous. Indeed, in a laboratory experiments with a one-shot prisoner’s dilemma (PD) and perfect stranger anonymity subjects (n=174) showed significantly different cooperation rates depending on the anonymity conditions during assembling and the payment procedure, ranging from 33.3% to 19.9%. In addition, a first experiment with the PD and anonymous payment and double blind experimenting lead to a cooperation rate of 33.3%. Only after the same subjects (n=34) took part a second time in same, entire experiment, the cooperation rate fell to 8.8%. Therefore this measurement of the cooperation rates in a laboratory experiment failed the test-retest check on reliability. This happened though all manipulations used fulfilled the standards of fully anonymous experimenting. This means that such processes could go unnoticed and bias the results of any standard laboratory experiment on cooperation in one shot decisions. Therefore, in accordance with the textbook logic of laboratory experiments, but in contrast to a common practice (cf. Behavioral Game Theory) marginal totals from cooperation experiments in the laboratory should not be interpreted.

Page generated in 0.0731 seconds