Abundance of labelled data played a crucial role in the recent developments in computer vision, but that faces problems like scalability and transferability to the wild. One alternative approach is to utilize the data without labels, i.e. unsupervised learning, in learning valuable information and put it in use to tackle vision problems. Generative Adversarial Networks (GANs) have gained momentum for their ability to model image distributions in unsupervised manner. They learn to emulate the training set and that enables sampling from that domain and using the knowledge learned for useful applications. Several methods proposed enhancing GANs, including regularizing the loss with some feature matching. We seek to push GANs beyond the data in the training and try to explore unseen territory in the image manifold. We first propose a new regularizer for GAN based on K-Nearest Neighbor (K-NN) selective feature matching to a target set Y in high-level feature space, during the adversarial training of GAN on the base set X, and we call this novel model K-GAN. We show that minimizing the added term follows from cross-entropy minimization between the distributions of GAN and set Y. Then, we introduce a cascaded framework for GANs that try to address the task of imagining a new distribution that combines the base set X and target set Y by cascading sampling GANs with translation GANs, and we dub the cascade of such GANs as the Imaginative Adversarial Network (IAN). Several cascades are trained on a collected dataset Zoo-Faces and generated innovative samples are shown, including from K-GAN cascade. We conduct an objective and subjective evaluation for different IAN setups in the addressed task of generating innovative samples and we show the effect of regularizing GAN on different scores. We conclude with some useful applications for these IANs, like multi-domain manifold traversing.

The main goal of this paper is the identification of comments on internet websites. The theoretical part is focused on artificial intelligence, mainly classifiers are described there. The practical part deals with creation of training database, which is formed by using generators of features. A generated feature might be for example a title of the HTML element where the comment is. The training database is created by input of classifiers. The result of this paper is testing classifiers in the RapidMiner program.

This paper describes basic methods used for optical character recognition. It explains all procedures of recognition from adjustment of picture, processing, feature extracting to matching algorithms. It compares methods and algorithms for character recognition obtained graphically distorted or else modified image so-called „captcha“, used in present. Further it compares method based on invariant moments and neural network as final classifier and method based on correlation between normals and recognized characters.

<div>Federated learning (FL) is a decentralized privacy-preserving learning technique in which clients learn a joint collaborative model through a central aggregator without sharing their data. In this setting, all clients learn a single common predictor (FedAvg), which does not generalize well on each client's local data due to the statistical data heterogeneity among clients. In this paper, we address this problem with PersFL, a discrete two-stage personalized learning algorithm. In the first stage, PersFL finds the optimal teacher model of each client during the FL training phase. In the second stage, PersFL distills the useful knowledge from optimal teachers into each user's local model. The teacher model provides each client with some rich, high-level representation that a client can easily adapt to its local model, which overcomes the statistical heterogeneity present at different clients. We evaluate PersFL on CIFAR-10 and MNIST datasets using three data-splitting strategies to control the diversity between clients' data distributions.</div><div><br></div><div>We empirically show that PersFL outperforms FedAvg and three state-of-the-art personalization methods, pFedMe, Per-FedAvg and FedPer on majority data-splits with minimal communication cost. Further, we study the performance of PersFL on different distillation objectives, how this performance is affected by the equitable notion of fairness among clients, and the number of required communication rounds. We also build an evaluation framework with the following modules: Data Generator, Federated Model Generation, and Evaluation Metrics. We introduce new metrics for the domain of personalized FL, and split these metrics into two perspectives: Performance, and Fairness. We analyze the performance of all the personalized algorithms by applying these metrics to answer the following questions: Which personalization algorithm performs the best in terms of accuracy across all the users?, and Which personalization algorithm is the fairest amongst all of them? Finally, we make the code for this work available at https://tinyurl.com/1hp9ywfa for public use and validation.</div>

No description available.

Developing high quality software is the goal of every software development organization. Software quality models are commonly used to assess and improve the software quality. These models, based on the past releases of the system, can be used to identify the fault-prone modules for the next release. This information is useful to the open-source software community, including both developers and users. Developers can use this information to clean or rebuild the faulty modules thus enhancing the system. The users of the software system can make informed decisions about the quality of the product. This thesis builds quality models using logistic regression, neural networks, decision trees, and genetic algorithms and compares their performance. Our results show that an overall accuracy of 65 ? 85% is achieved with a type II misclassification rate of approximately 20 ? 35%. Performance of each of the methods is comparable to the others with minor variations.

No description available.

No description available.

This master thesis deals with classification of sleep stages on the base of polysomnographic signals. On several signals was performed analysis and feature extraxtion in time domain and in frequency domain as well. For feature extraxtion was used EEG, EOG and EMG signals. For classification was selected classification models K-NN, SVM and artifical neural network. Accuracy of classifation is different depending on used method and spleep stages split. The best results achieved classification among stages Wake, REM, and N3, with neural network usage. In this case the succes was 93,1 %.

Made available in DSpace on 2015-04-22T22:06:14Z (GMT). No. of bitstreams: 1 Fernando Herkrath.pdf: 1327969 bytes, checksum: 4edb24ec4019887e8aec0015708ef00b (MD5) Previous issue date: 2011-06-17 / The normative needs (NN) professionally defined have been widely used for recommending orthodontic treatment. The socio-dental approach for treatment needs is a method that embodies to NN self-reported subjective measures and oral health related behaviors. It was proposed because of the limitations of the NN methods. The aim of this study was to test the socio-dental approach of orthodontic treatment needs in adolescents. A cross-sectional study was conducted in the city of Manaus, Amazonas, Brazil in 2010-2011 and included 201 scholars of 12 years-old. The orthodontic treatment NN was determined by oral clinical examination using two occlusal indexes, Index of Orthodontic Treatment Need (IOTN) and Dental Aesthetics Index (DAI). The socio-dental approach has combined normative measures (IOTN and DAI), impact of malocclusion on daily activities (Child-OIDP), and propensityrelated orthodontic treatment assessment. The methods of treatment needs assessment were compared using McNemar test. The association between the impact of malocclusion on daily activities and NN was tested by Qui-square test and Kruskal-Wallis test. The frequency of individuals with NN treatment (IOTN and DAI) was statistically higher comparing to sociodental approach (p<0.001). The socio-dental approach showed different results when using IOTN and DAI (p<0.001). The magnitude of NN was associated with the impact of malocclusion on the adolescents daily activities. Large reductions in normative needs estimates for orthodontic treatment were apparent using socio-dental approach. The sociodental approach for orthodontic treatment needs can optimize the use of resources in oral health services. / A necessidade normativa (NN) definida pelo profissional têm sido amplamente utilizada para indicação de tratamento ortodôntico. Devido às limitações deste método foi proposta a avaliação sociodental de necessidade de tratamento, um método que incorpora à NN medidas subjetivas autorreferidas e comportamentos relacionados à saúde bucal. O objetivo deste estudo foi testar a avaliação sociodental de necessidade de tratamento ortodôntico em adolescentes. Um estudo seccional envolveu 201 escolares de 12 anos de idade na cidade de Manaus, Amazonas, em 2010-2011. A NN de tratamento ortodôntico foi determinada através de exame clínico bucal e o emprego do Índice de Necessidade de Tratamento Ortodôntico (IOTN) e do Índice de Estética Dental (DAI). A abordagem sociodental combinou medidas normativas (IOTN e DAI), impacto da má oclusão nas atividades diárias (instrumento Child- OIDP) e avaliação da propensão ao tratamento ortodôntico. Os métodos de avaliação de necessidade de tratamento foram comparados entre si com o teste de McNemar. A associação entre o impacto da má oclusão nas atividades diárias e a NN foi testada com o teste Quiquadrado e teste de Kruskal-Wallis. A frequência de indivíduos com NN de tratamento (IOTN e DAI) foi estatisticamente maior em comparação com a abordagem sociodental (p<0,001). A abordagem sociodental mostrou resultados distintos utilizando-se o DAI e o IOTN (p<0,001). A magnitude da NN foi associada ao impacto da má oclusão nas atividades diárias dos adolescentes. Observou-se uma grande redução nas estimativas de necessidade de tratamento ortodôntico em adolescentes com o emprego da abordagem sociodental. O uso da avaliação sociodental de necessidade de tratamento ortodôntico pode otimizar o emprego dos recursos nos serviços públicos odontológicos.