Goal of this work is to make acquaintance and study meta-learningu methods, program algorithm and compare with other machine learning methods.

This thesis is dedicated to the principles of a MP3 audio format decoding on available develpoment kits powered by ARM processors. It compares kit MCB2300 with LPC2378 microcontroller and kit Raspberry Pi with Debian operation system. The comparsion is focused on their suitability for MP3 decoder implementation. A complete design of support hardware modules, which complement the missing hardware of development kits is described in detail. Thesis includes the realization of an implementation of a MP3 decoder with additional visual effects in a form of an attached RGB LED strip developed on a Raspberry Pi development kit.

Uncontrolled growth of vascular stem cells as a result of endothelial-mesenchymal transition is considered to cause hyper-proliferative vascular remodeling in severe pulmonary arterial hypertension (PAH) patients. Hyperplastic intimal growth is one of the causes of closure of the lumen of pulmonary arterioles. This abnormal vessel remodeling leads to the progressive increase in pressure of the pulmonary arterioles causing severe PAH; and debilitating harm to patients resulting in mortality from right heart failure. Environmental factors, including polychlorinated biphenyls (PCBs), are considered to be involved in hyper-proliferative vascular remodeling because genetic makeup can only explain about 10% of severe PAH cases. PCB involvement in lung toxicity has received attention because (i) they have been reported to accumulate in the lung; (ii) PCBs produce pathological vascular remodeling in the experimental model; high levels of PCBs are found in human lung tissue; and (iii) epidemiological studies show the association between lung toxicity and PCBs; and prevalence of hypertension and elevated concentrations of particularly PCB153. Recent studies identify PCB153 as one of the largest contributors for total PCB body burden in humans. Our previous studies demonstrated PCB153 mediated vascular endothelial dysfunction and activated the inhibitor of differentiation protein 3 (ID3). ID3 is an important determinant of mitogen and reactive oxygen species-induced G1→S phase cell cycle progression. Although phosphorylation of ID3 increases cell growth by antagonizing the transcription of cell cycle inhibitors, still there is a critical gap in understanding the molecular mechanism(s) of pulmonary proliferative vascular remodeling associated with PCB exposure in humans and the role of the transcription regulator ID3. Our overall objective was to investigate ID3 mediated transcriptional reprogramming as a driver of PCB153-induced pathological proliferative vascular remodeling. Stable ectopic expression of ID3 in lung endothelial cells contributed to endothelial-mesenchymal transition (EndMT), cell proliferation, and cell migration. Using an endothelial spheroid assay, an established method to measure aberrant hyper-proliferation of endothelial cells in PAH patients, we show that stable ectopic expression of ID3 increased the number and size of vascular spheres. ID3 overexpressing cells exposed to environmentally relevant concentrations of PCB153 showed a two-fold increase in cell proliferation as determined by MTT, SRB, and BrdU assays. ID3 overexpressing cells showed the loss of VE-cadherin and gain of MMP9 and vimentin, which are markers of EndMT. PCB153 also increased phosphorylation of ID3 in lung endothelial cells. To determine the molecular mechanism by which ID3 contributes to hyper-proliferative endothelial cells, we investigated ID3 transcriptional reprogramming using ChIP-Seq and RNA-Seq technology. We show here for the first time that ID3 is part of a more general mechanism of transcriptional regulation. Our ChIP-Seq data show that ID3 binds to a subset of approximately 1200 target genes. Comprehensive motif analysis of ChIP-Seq data using the MEME Suite software toolkit revealed that ID3 bound to the GAGAGAGAGA motif sequence on genomic DNA. We also show a significant preference of ID3 binding to motifs associated with transcription factors IRF1, BC11A, IRF4, PRDM1, FOXJ3, SMAD4, ZBTB6, GATA1, and STAT2. Using an integrative approach of ChIP-Seq and RNA-Seq data, we identified 19 genes whose promoter region was bound by ID3 and RNA was differentially expressed in ID3 overexpressing cells. In summary, our data demonstrated that PCB153 and/or ID3 induces proliferation of lung endothelial cells via transcriptional reprogramming. Discoveries from these findings will lay the necessary groundbreaking work for testing the efficacy of ID3 antagonists for the prevention and treatment of pathological vascular remodeling as well as provide a new paradigm by which PCBs may contribute to lung vascular toxicity.

Privacy preservation is important for machine learning and datamining, but measures designed to protect private information sometimes result in a trade off: reduced utility of the training samples. This thesis introduces a privacy preserving approach that can be applied to decision-tree learning, without concomitant loss of accuracy. It describes an approach to the preservation of privacy of collected data samples in cases when information of the sample database has been partially lost. This approach converts the original sample datasets into a group of unreal datasets, where an original sample cannot be reconstructed without the entire group of unreal datasets. This approach does not perform well for sample datasets with low frequency, or when there is low variance in the distribution of all samples. However, this problem can be solved through a modified implementation of the approach introduced later in this thesis, by using some extra storage.

This thesis discusses the possibility of combining music and images through the use of metadata. Test subjects from different usability tests say they are interested in seeing images of the band or artist they are listening too. Lyrics matching the actual song are also something they would like to see. As a result an application for cellphones is created with Flash Lite which shows that it is possible to listen to music and automatically get images from Flickr and lyrics from Lyrictracker which match the music and show them on a cellphone.

La crête neurale (CN) est une structure transitoire apparaissant en bordure de la plaque neurale chez les embryons de vertébrés. Au cours du développement embryonnaire, les cellules de la CN prolifèrent, subissent une transition épithélio-mésenchymateuse, migrent et se différencient en de nombreux types cellulaires tels que des neurones et cellules gliales du système nerveux périphérique, des mélanocytes, des cellules musculaires lisses ou des élements du squelette cranio-facial. Afin de mieux comprendre les mécanismes moléculaires contrôlant la prolifération et la spécification des cellules de la CN, nous avons étudié le rôle de deux facteurs de transcription, Hairy2 et Stat3, via des expériences de perte et gain de fonction chez l’embryon de xénope. <p>Le gène Hairy2 code pour un facteur de transcription bHLH-O répresseur. Il est exprimé précocement au niveau de la bordure de la plaque neurale incluant la CN présomptive. Nous avons montré que Hairy2 est requis pour la prolifération des cellules de la CN en aval de signaux FGFs et qu’il maintient les cellules dans un état indifférencié en réprimant l’expression précoce des gènes spécifiques de la CN. Hairy2 réprime aussi la transcription du gène Id3 codant pour un facteur HLH essentiel à la prolifération des cellules de la CN. Id3 affecte également Hairy2. Nous avons observé que la protéine Id3 interagit physiquement avec Hairy2 et bloque son activité, démontrant que les interactions entre Hairy2 et Id3 jouent un rôle important dans la prolifération et la spécification des cellules de la CN. <p>\ / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

The work deals with verification of the metadata of audio recordings in an environment of mobile phones running Android. The text work is divided into three main parts. The first part is focused more on theoretical basis, where the main emphasis on the description of Android, application structure, description of the audio MP3 file and theoretical basis of Fourier transform in signal processing. The second part is devoted to solution design, analysis and implementation. The third part deals with testing.

The main goal of this master's thesis was to study the main principles of classification methods. Basic principles of knowledge discovery process, data mining and using an external class CSSBox are described. Special attantion was paid to implementation of a ,,k-nearest neighbors`` classification method. The first objective of this work was to create training and testing data described by 'n' attributes. The second objective was to perform experimental analysis to determine a good value for 'k', the number of neighbors.