Improving sampling, optimization and feature extraction in Boltzmann machines

L’apprentissage supervisé de réseaux hiérarchiques à grande échelle connaît présentement un succès fulgurant. Malgré cette effervescence, l’apprentissage non-supervisé représente toujours, selon plusieurs chercheurs, un élément clé de l’Intelligence Artificielle, où les agents doivent apprendre à partir d’un nombre potentiellement limité de données. Cette thèse s’inscrit dans cette pensée et aborde divers sujets de recherche liés au problème d’estimation de densité par l’entremise des machines de Boltzmann (BM), modèles graphiques probabilistes au coeur de l’apprentissage profond. Nos contributions touchent les domaines de l’échantillonnage, l’estimation de fonctions de partition, l’optimisation ainsi que l’apprentissage de représentations invariantes. Cette thèse débute par l’exposition d’un nouvel algorithme d'échantillonnage adaptatif, qui ajuste (de fa ̧con automatique) la température des chaînes de Markov sous simulation, afin de maintenir une vitesse de convergence élevée tout au long de l’apprentissage. Lorsqu’utilisé dans le contexte de l’apprentissage par maximum de vraisemblance stochastique (SML), notre algorithme engendre une robustesse accrue face à la sélection du taux d’apprentissage, ainsi qu’une meilleure vitesse de convergence. Nos résultats sont présent ́es dans le domaine des BMs, mais la méthode est générale et applicable à l’apprentissage de tout modèle probabiliste exploitant l’échantillonnage par chaînes de Markov. Tandis que le gradient du maximum de vraisemblance peut-être approximé par échantillonnage, l’évaluation de la log-vraisemblance nécessite un estimé de la fonction de partition. Contrairement aux approches traditionnelles qui considèrent un modèle donné comme une boîte noire, nous proposons plutôt d’exploiter la dynamique de l’apprentissage en estimant les changements successifs de log-partition encourus à chaque mise à jour des paramètres. Le problème d’estimation est reformulé comme un problème d’inférence similaire au filtre de Kalman, mais sur un graphe bi-dimensionnel, où les dimensions correspondent aux axes du temps et au paramètre de température. Sur le thème de l’optimisation, nous présentons également un algorithme permettant d’appliquer, de manière efficace, le gradient naturel à des machines de Boltzmann comportant des milliers d’unités. Jusqu’à présent, son adoption était limitée par son haut coût computationel ainsi que sa demande en mémoire. Notre algorithme, Metric-Free Natural Gradient (MFNG), permet d’éviter le calcul explicite de la matrice d’information de Fisher (et son inverse) en exploitant un solveur linéaire combiné à un produit matrice-vecteur efficace. L’algorithme est prometteur: en terme du nombre d’évaluations de fonctions, MFNG converge plus rapidement que SML. Son implémentation demeure malheureusement inefficace en temps de calcul. Ces travaux explorent également les mécanismes sous-jacents à l’apprentissage de représentations invariantes. À cette fin, nous utilisons la famille de machines de Boltzmann restreintes “spike & slab” (ssRBM), que nous modifions afin de pouvoir modéliser des distributions binaires et parcimonieuses. Les variables latentes binaires de la ssRBM peuvent être rendues invariantes à un sous-espace vectoriel, en associant à chacune d’elles, un vecteur de variables latentes continues (dénommées “slabs”). Ceci se traduit par une invariance accrue au niveau de la représentation et un meilleur taux de classification lorsque peu de données étiquetées sont disponibles. Nous terminons cette thèse sur un sujet ambitieux: l’apprentissage de représentations pouvant séparer les facteurs de variations présents dans le signal d’entrée. Nous proposons une solution à base de ssRBM bilinéaire (avec deux groupes de facteurs latents) et formulons le problème comme l’un de “pooling” dans des sous-espaces vectoriels complémentaires.

Limited sampling strategies for estimation of cyclosporine exposure in pediatric hematopoietic stem cell transplant recipients : methodological improvement and introduction of sampling time deviation analysis.

Le suivi thérapeutique est recommandé pour l’ajustement de la dose des agents immunosuppresseurs. La pertinence de l’utilisation de la surface sous la courbe (SSC) comme biomarqueur dans l’exercice du suivi thérapeutique de la cyclosporine (CsA) dans la transplantation des cellules souches hématopoïétiques est soutenue par un nombre croissant d’études. Cependant, pour des raisons intrinsèques à la méthode de calcul de la SSC, son utilisation en milieu clinique n’est pas pratique. Les stratégies d’échantillonnage limitées, basées sur des approches de régression (R-LSS) ou des approches Bayésiennes (B-LSS), représentent des alternatives pratiques pour une estimation satisfaisante de la SSC. Cependant, pour une application efficace de ces méthodologies, leur conception doit accommoder la réalité clinique, notamment en requérant un nombre minimal de concentrations échelonnées sur une courte durée d’échantillonnage. De plus, une attention particulière devrait être accordée à assurer leur développement et validation adéquates. Il est aussi important de mentionner que l’irrégularité dans le temps de la collecte des échantillons sanguins peut avoir un impact non-négligeable sur la performance prédictive des R-LSS. Or, à ce jour, cet impact n’a fait l’objet d’aucune étude. Cette thèse de doctorat se penche sur ces problématiques afin de permettre une estimation précise et pratique de la SSC. Ces études ont été effectuées dans le cadre de l’utilisation de la CsA chez des patients pédiatriques ayant subi une greffe de cellules souches hématopoïétiques. D’abord, des approches de régression multiple ainsi que d’analyse pharmacocinétique de population (Pop-PK) ont été utilisées de façon constructive afin de développer et de valider adéquatement des LSS. Ensuite, plusieurs modèles Pop-PK ont été évalués, tout en gardant à l’esprit leur utilisation prévue dans le contexte de l’estimation de la SSC. Aussi, la performance des B-LSS ciblant différentes versions de SSC a également été étudiée. Enfin, l’impact des écarts entre les temps d’échantillonnage sanguins réels et les temps nominaux planifiés, sur la performance de prédiction des R-LSS a été quantifié en utilisant une approche de simulation qui considère des scénarios diversifiés et réalistes représentant des erreurs potentielles dans la cédule des échantillons sanguins. Ainsi, cette étude a d’abord conduit au développement de R-LSS et B-LSS ayant une performance clinique satisfaisante, et qui sont pratiques puisqu’elles impliquent 4 points d’échantillonnage ou moins obtenus dans les 4 heures post-dose. Une fois l’analyse Pop-PK effectuée, un modèle structural à deux compartiments avec un temps de délai a été retenu. Cependant, le modèle final - notamment avec covariables - n’a pas amélioré la performance des B-LSS comparativement aux modèles structuraux (sans covariables). En outre, nous avons démontré que les B-LSS exhibent une meilleure performance pour la SSC dérivée des concentrations simulées qui excluent les erreurs résiduelles, que nous avons nommée « underlying AUC », comparée à la SSC observée qui est directement calculée à partir des concentrations mesurées. Enfin, nos résultats ont prouvé que l’irrégularité des temps de la collecte des échantillons sanguins a un impact important sur la performance prédictive des R-LSS; cet impact est en fonction du nombre des échantillons requis, mais encore davantage en fonction de la durée du processus d’échantillonnage impliqué. Nous avons aussi mis en évidence que les erreurs d’échantillonnage commises aux moments où la concentration change rapidement sont celles qui affectent le plus le pouvoir prédictif des R-LSS. Plus intéressant, nous avons mis en exergue que même si différentes R-LSS peuvent avoir des performances similaires lorsque basées sur des temps nominaux, leurs tolérances aux erreurs des temps d’échantillonnage peuvent largement différer. En fait, une considération adéquate de l'impact de ces erreurs peut conduire à une sélection et une utilisation plus fiables des R-LSS. Par une investigation approfondie de différents aspects sous-jacents aux stratégies d’échantillonnages limités, cette thèse a pu fournir des améliorations méthodologiques notables, et proposer de nouvelles voies pour assurer leur utilisation de façon fiable et informée, tout en favorisant leur adéquation à la pratique clinique.

Whither RDS? An investigation of respondent driven sampling as a method of recruiting mainstream marijuana users.

Background: An important challenge in conducting social research of specific relevance to harm reduction programs is locating hidden populations of consumers of substances like cannabis who typically report few adverse or unwanted consequences of their use. Much of the deviant, pathologized perception of drug users is historically derived from, and empirically supported, by a research emphasis on gaining ready access to users in drug treatment or in prison populations with higher incidence of problems of dependence and misuse. Because they are less visible, responsible recreational users of illicit drugs have been more difficult to study. Methods: This article investigates Respondent Driven Sampling (RDS) as a method of recruiting experienced marijuana users representative of users in the general population. Based on sampling conducted in a multi-city study (Halifax, Montreal, Toronto, and Vancouver), and compared to samples gathered using other research methods, we assess the strengths and weaknesses of RDS recruitment as a means of gaining access to illicit substance users who experience few harmful consequences of their use. Demographic characteristics of the sample in Toronto are compared with those of users in a recent household survey and a pilot study of Toronto where the latter utilized nonrandom self-selection of respondents. Results: A modified approach to RDS was necessary to attain the target sample size in all four cities (i.e., 40 'users' from each site). The final sample in Toronto was largely similar, however, to marijuana users in a random household survey that was carried out in the same city. Whereas well-educated, married, whites and females in the survey were all somewhat overrepresented, the two samples, overall, were more alike than different with respect to economic status and employment. Furthermore, comparison with a self-selected sample suggests that (even modified) RDS recruitment is a cost-effective way of gathering respondents who are more representative of users in the general population than nonrandom methods of recruitment ordinarily produce. Conclusions: Research on marijuana use, and other forms of drug use hidden in the general population of adults, is important for informing and extending harm reduction beyond its current emphasis on 'at-risk' populations. Expanding harm reduction in a normalizing context, through innovative research on users often overlooked, further challenges assumptions about reducing harm through prohibition of drug use and urges consideration of alternative policies such as decriminalization and legal regulation.

Chronic catheterization using vascular-access-port in rats: blood sampling with minimal stress for plasma catecholamine determination

Chronic catheterization is illustrated using vascular-access-port model SLA where the port is surgically placed subcutaneously on the back of the rat. The catheter is tunnelled to the neck and inserted into the jugular vein . Within 24 h rats showed normal blood pressure and blood samples were collected at intervals with minimal stress to the animals . A comparison of the plasma catecholamine of blood collected from vascular-access-ports with that obtained from decapitation indicates that there was minimal stress to the rats when blood was drawn through the vascular-access-port.

Secret Sharing Schemes Using Visual Cryptography

Secret sharing schemes allow a secret to be shared among a group of participants so that only qualified subsets of participants can recover the secret. A visual cryptography scheme (VCS) is a special kind of secret sharing scheme in which the secret to share consists of an image and the shares consist of xeroxed transparencies which are stacked to recover the shared image. In this thesis we have given the theoretical background of Secret Sharing Schemes and the historical development of the subject. We have included a few examples to improve the readability of the thesis. We have tried to maintain the rigor of the treatment of the subject. The limitations and disadvantages of the various forms secret sharing schemes are brought out. Several new schemes for both dealing and combining are included in the thesis. We have introduced a new number system, called, POB number system. Representation using POB number system has been presented. Algorithms for finding the POB number and POB value are given.We have also proved that the representation using POB number system is unique and is more efficient. Being a new system, there is much scope for further development in this area.

Energy-Efficient Cluster based Routing Schemes for Static and Mobile Sensor Networks

The theme of the thesis is centred around one important aspect of wireless sensor networks; the energy-efficiency.The limited energy source of the sensor nodes calls for design of energy-efficient routing protocols. The schemes for protocol design should try to minimize the number of communications among the nodes to save energy. Cluster based techniques were found energy-efficient. In this method clusters are formed and data from different nodes are collected under a cluster head belonging to each clusters and then forwarded it to the base station.Appropriate cluster head selection process and generation of desirable distribution of the clusters can reduce energy consumption of the network and prolong the network lifetime. In this work two such schemes were developed for static wireless sensor networks.In the first scheme, the energy wastage due to cluster rebuilding incorporating all the nodes were addressed. A tree based scheme is presented to alleviate this problem by rebuilding only sub clusters of the network. An analytical model of energy consumption of proposed scheme is developed and the scheme is compared with existing cluster based scheme. The simulation study proved the energy savings observed.The second scheme concentrated to build load-balanced energy efficient clusters to prolong the lifetime of the network. A voting based approach to utilise the neighbor node information in the cluster head selection process is proposed. The number of nodes joining a cluster is restricted to have equal sized optimum clusters. Multi-hop communication among the cluster heads is also introduced to reduce the energy consumption. The simulation study has shown that the scheme results in balanced clusters and the network achieves reduction in energy consumption.The main conclusion from the study was the routing scheme should pay attention on successful data delivery from node to base station in addition to the energy-efficiency. The cluster based protocols are extended from static scenario to mobile scenario by various authors. None of the proposals addresses cluster head election appropriately in view of mobility. An elegant scheme for electing cluster heads is presented to meet the challenge of handling cluster durability when all the nodes in the network are moving. The scheme has been simulated and compared with a similar approach.The proliferation of sensor networks enables users with large set of sensor information to utilise them in various applications. The sensor network programming is inherently difficult due to various reasons. There must be an elegant way to collect the data gathered by sensor networks with out worrying about the underlying structure of the network. The final work presented addresses a way to collect data from a sensor network and present it to the users in a flexible way.A service oriented architecture based application is built and data collection task is presented as a web service. This will enable composition of sensor data from different sensor networks to build interesting applications. The main objective of the thesis was to design energy-efficient routing schemes for both static as well as mobile sensor networks. A progressive approach was followed to achieve this goal.

Implications of Government Sponsored Social Security and Welfare Schemes on Socioeconomic Status of Small Scale Marine Fisherfolk in Kerala

A socio-economic research is required as an attempt to address the socio-economic issues facing small-scale fisheries. A study of the socio economic conditions of small-scale fishermen is a prerequisite for good design and successful implementation of effective assistance Programmes. It will provide an overall pidure of the structure, activities and standards of living of small-scale fisherfolk The study is confined to the coastal districts of Ernakulam, Thrissur and Malappuram districts. It also gives a picture of socio-economic conditions of the fisher folk in the study area. The variables that may depict the standard of living of the small-scale fisherfolk are occupational structure, family size, age structure, income, expenditure, education, housing and other social amenities. It attempts to see the asset creation of the fisherfolk with the help of government agencies, and the nature of savings and expenditure pattern of the fisherfolk. It also provides a picture of the indebtedness of the fisherfolk in the study area. The study analyses the schemes implemented by the government through its agencies, like Fisheries Department, Matsyaboard, and Matsyafed; and the awareness of fisherfolk regarding these schemes, their attitude and reactions, the extent of accessibility, and the viability of the schemes.

Environomics - A Financial Estimate of Environmental Pollution Control and Abatement Schemes in Eloor-Edayar Industrial Belt

The present study consists of nine chapters including the introductory chapter. Chapter II makes a brief review of environmental literature and examines various measures adopted at the global level to protect the environment. The environmental problems often transgress national sovereignity and geographical boundaries. Therefore, attempts must be made at the national and international levels to protect the environment, the resources of which are the common property of mankind. The protection of the national environment from the ancient till the present forms the content of Chapter III. These chapters together provide a background to understand the issues analysed in the subsequent chapters. Carefully worked out theoretical framework is a pre-requisite for the successful study of a complex subject. Some of the theoretical issues of ‘environomics’ are examined in Chapter IV. The theoretical issues involved in estimating the costs and benefits of environmental protection constitute the theme of Chapter V. The state of environment in Eloor-Edayar Industrial belt andthe impact analysis of pollution of the area are discussed in Chapter VI and VII respectively. Chapter VIII makes the financial estimate of environmental protection of the project And finally, Chapter IX presents the findings of the study

A Comparative Study of Convective Parameterization Schemes in WRF-NMM Model

Severe local storms, including tornadoes, damaging hail and wind gusts, frequently occur over the eastern and northeastern states of India during the pre-monsoon season (March-May). Forecasting thunderstorms is one of the most difficult tasks in weather prediction, due to their rather small spatial and temporal extension and the inherent non-linearity of their dynamics and physics. In this paper, sensitivity experiments are conducted with the WRF-NMM model to test the impact of convective parameterization schemes on simulating severe thunderstorms that occurred over Kolkata on 20 May 2006 and 21 May 2007 and validated the model results with observation. In addition, a simulation without convective parameterization scheme was performed for each case to determine if the model could simulate the convection explicitly. A statistical analysis based on mean absolute error, root mean square error and correlation coefficient is performed for comparisons between the simulated and observed data with different convective schemes. This study shows that the prediction of thunderstorm affected parameters is sensitive to convective schemes. The Grell-Devenyi cloud ensemble convective scheme is well simulated the thunderstorm activities in terms of time, intensity and the region of occurrence of the events as compared to other convective schemes and also explicit scheme

Extension schemes for the Alignment Model of English-Malayalam Statistical Machine Translator

In Statistical Machine Translation from English to Malayalam, an unseen English sentence is translated into its equivalent Malayalam sentence using statistical models. A parallel corpus of English-Malayalam is used in the training phase. Word to word alignments has to be set among the sentence pairs of the source and target language before subjecting them for training. This paper deals with certain techniques which can be adopted for improving the alignment model of SMT. Methods to incorporate the parts of speech information into the bilingual corpus has resulted in eliminating many of the insignificant alignments. Also identifying the name entities and cognates present in the sentence pairs has proved to be advantageous while setting up the alignments. Presence of Malayalam words with predictable translations has also contributed in reducing the insignificant alignments. Moreover, reduction of the unwanted alignments has brought in better training results. Experiments conducted on a sample corpus have generated reasonably good Malayalam translations and the results are verified with F measure, BLEU and WER evaluation metrics.

Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation

Presently different audio watermarking methods are available; most of them inclined towards copyright protection and copy protection. This is the key motive for the notion to develop a speaker verification scheme that guar- antees non-repudiation services and the thesis is its outcome. The research presented in this thesis scrutinizes the field of audio water- marking and the outcome is a speaker verification scheme that is proficient in addressing issues allied to non-repudiation to a great extent. This work aimed in developing novel audio watermarking schemes utilizing the fun- damental ideas of Fast-Fourier Transform (FFT) or Fast Walsh-Hadamard Transform (FWHT). The Mel-Frequency Cepstral Coefficients (MFCC) the best parametric representation of the acoustic signals along with few other key acoustic characteristics is employed in crafting of new schemes. The au- dio watermark created is entirely dependent to the acoustic features, hence named as FeatureMark and is crucial in this work. In any watermarking scheme, the quality of the extracted watermark de- pends exclusively on the pre-processing action and in this work framing and windowing techniques are involved. The theme non-repudiation provides immense significance in the audio watermarking schemes proposed in this work. Modification of the signal spectrum is achieved in a variety of ways by selecting appropriate FFT/FWHT coefficients and the watermarking schemes were evaluated for imperceptibility, robustness and capacity char- acteristics. The proposed schemes are unequivocally effective in terms of maintaining the sound quality, retrieving the embedded FeatureMark and in terms of the capacity to hold the mark bits. Robust nature of these marking schemes is achieved with the help of syn- chronization codes such as Barker Code with FFT based FeatureMarking scheme and Walsh Code with FWHT based FeatureMarking scheme. An- other important feature associated with this scheme is the employment of an encryption scheme towards the preparation of its FeatureMark that scrambles the signal features that helps to keep the signal features unreve- laed. A comparative study with the existing watermarking schemes and the ex- periments to evaluate imperceptibility, robustness and capacity tests guar- antee that the proposed schemes can be baselined as efficient audio water- marking schemes. The four new digital audio watermarking algorithms in terms of their performance are remarkable thereby opening more opportu- nities for further research.

Stability of preconditioned finite volume schemes at low Mach numbers

In [4], Guillard and Viozat propose a finite volume method for the simulation of inviscid steady as well as unsteady flows at low Mach numbers, based on a preconditioning technique. The scheme satisfies the results of a single scale asymptotic analysis in a discrete sense and comprises the advantage that this can be derived by a slight modification of the dissipation term within the numerical flux function. Unfortunately, it can be observed by numerical experiments that the preconditioned approach combined with an explicit time integration scheme turns out to be unstable if the time step Dt does not satisfy the requirement to be O(M2) as the Mach number M tends to zero, whereas the corresponding standard method remains stable up to Dt=O(M), M to 0, which results from the well-known CFL-condition. We present a comprehensive mathematical substantiation of this numerical phenomenon by means of a von Neumann stability analysis, which reveals that in contrast to the standard approach, the dissipation matrix of the preconditioned numerical flux function possesses an eigenvalue growing like M-2 as M tends to zero, thus causing the diminishment of the stability region of the explicit scheme. Thereby, we present statements for both the standard preconditioner used by Guillard and Viozat [4] and the more general one due to Turkel [21]. The theoretical results are after wards confirmed by numerical experiments.

Numerical simulation of tunnel fires using preconditioned finite volume schemes

This article is concerned with the numerical simulation of flows at low Mach numbers which are subject to the gravitational force and strong heat sources. As a specific example for such flows, a fire event in a car tunnel will be considered in detail. The low Mach flow is treated with a preconditioning technique allowing the computation of unsteady flows, while the source terms for gravitation and heat are incorporated via operator splitting. It is shown that a first order discretization in space is not able to compute the buoyancy forces properly on reasonable grids. The feasibility of the method is demonstrated on several test cases.

Labelling schemes for organic food: Consumer attitudes, preferences and willingness-to-pay in six European countries

Since July 2010, all pre-packed organic food produced in the European Union (EU) must carry the new mandatory EU logo for organic food. Voluntary organic labels (such as national governmental logos and logos of private farmers’ associations and control bodies) can still be used, but only in addition to the mandatory EU logo. This change in the regulatory environment of organic labelling raises a number of questions regarding consumer preferences for different kinds of organic certification logos, which the present dissertation addressed. The first objective was to explore and analyse consumer perceptions, attitudes, preferences and willingness-to-pay (WTP) regarding different voluntary organic labels. The second objective was to investigate consumer perceptions and attitudes towards a mandatory EU logo. A combination of qualitative and quantitative methods of consumer research in six European countries was employed including focus group discussions, choice experiments and structured interviews. Based on the empirical results, recommendations are drawn for different actors in the organic sector.