We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions about data, and view learning algorithms as methods for finding parameter values that look probable in the light of the data. We discuss basic issues in representation and learning, and treat some of the practical issues that arise in fitting networks to data. We also discuss links between neural networks and the general formalism of graphical models.

No description available.

One of the most important and challenging ``knowledge extraction' tasks in bioinformatics is the reverse engineering of gene regulatory networks (GRNs) from DNA microarray gene expression data. Indeed, as a result of the development of high-throughput data-collection techniques, biology is experiencing a data flood phenomenon that pushes biologists toward a new view of biology--systems biology--that aims at system-level understanding of biological systems. Unfortunately, even for small model organisms such as the yeast Saccharomyces cerevisiae, the number p of genes is much larger than the number n of expression data samples. The dimensionality issue induced by this ``small n, large p' data setting renders standard statistical learning methods inadequate. Restricting the complexity of the models enables to deal with this serious impediment. Indeed, by introducing (a priori undesirable) bias in the model selection procedure, one reduces the variance of the selected model thereby increasing its accuracy. Gaussian graphical models (GGMs) have proven to be a very powerful formalism to infer GRNs from expression data. Standard GGM selection techniques can unfortunately not be used in the ``small n, large p' data setting. One way to overcome this issue is to resort to regularization. In particular, shrinkage estimators of the covariance matrix--required to infer GGMs--have proven to be very effective. Our first contribution consists in a new shrinkage estimator that improves upon existing ones through the use of a Monte Carlo (parametric bootstrap) procedure. Another approach to GGM selection in the ``small n, large p' data setting consists in reverse engineering limited-order partial correlation graphs (q-partial correlation graphs) to approximate GGMs. Our second contribution consists in an inference algorithm, the q-nested procedure, that builds a sequence of nested q-partial correlation graphs to take advantage of the smaller order graphs' topology to infer higher order graphs. This allows us to significantly speed up the inference of such graphs and to avoid problems related to multiple testing. Consequently, we are able to consider higher order graphs, thereby increasing the accuracy of the inferred graphs. Another important challenge in bioinformatics is the prediction of gene function. An example of such a prediction task is the identification of genes that are targets of the nitrogen catabolite repression (NCR) selection mechanism in the yeast Saccharomyces cerevisiae. The study of model organisms such as Saccharomyces cerevisiae is indispensable for the understanding of more complex organisms. Our third contribution consists in extending the standard two-class classification approach by enriching the set of variables and comparing several feature selection techniques and classification algorithms. Finally, our fourth contribution formulates the prediction of NCR target genes as a network inference task. We use GGM selection to infer multivariate dependencies between genes, and, starting from a set of genes known to be sensitive to NCR, we classify the remaining genes. We hence avoid problems related to the choice of a negative training set and take advantage of the robustness of GGM selection techniques in the ``small n, large p' data setting.

There is a dilemma between security and usability, which are two fundamentally conflicting issues. From the usability perspective, authentication protocols should be easy to use and passwords generated from these protocols should be easy to remember. From the security perspective, passwords should be hard to guess and should not be written down or stored in a plain text. Instead of using text based passwords, graphical passwords have been proposed to increase both memorability and security. Biederman (1972) and Biederman, Glass, &amp / Stacy (1973) reported that the objects in a coherent image were recognized and identified more efficiently and quickly than the objects in a jumbled image in which the jumbled image was created by dividing the coherent image into sections and changing the position of the sections without rotating them. The study was designed to experimentally examine the differences in usability and security of the graphical password scheme by manipulating the coherence of the displayed image. Sixty-three volunteers participated in the main experiment. The participants were divided into groups according to the type of image they were presented in the password creation (either coherent-image or jumbled-image) task. Each participant created a graphical password and three days after the first session (i.e., second session) s/he tried to remember it in order to authenticate to the system. The results revealed that in the proposed graphical password scheme, using coherent image has more advantages over jumbled image in terms of usability and security.

Genetic data analysis has been capturing a lot of attentions for understanding the mechanism of the development and progressing of diseases like cancers, and is crucial in discovering genetic markers and treatment targets in medical research. This dissertation focuses on several important issues in genetic data analysis, graphical network modeling, feature selection, and covariance estimation. First, we develop a gene network modeling method for discrete gene expression data, produced by technologies such as serial analysis of gene expression and RNA sequencing experiment, which generate counts of mRNA transcripts in cell samples. We propose a generalized linear model to fit the discrete gene expression data and assume that the log ratios of the mean expression levels follow a Gaussian distribution. We derive the gene network structures by selecting covariance matrices of the Gaussian distribution with a hyper-inverse Wishart prior. We incorporate prior network models based on Gene Ontology information, which avails existing biological information on the genes of interest. Next, we consider a variable selection problem, where the variables have natural grouping structures, with application to analysis of chromosomal copy number data. The chromosomal copy number data are produced by molecular inversion probes experiments which measure probe-specific copy number changes. We propose a novel Bayesian variable selection method, the hierarchical structured variable se- lection (HSVS) method, which accounts for the natural gene and probe-within-gene architecture to identify important genes and probes associated with clinically relevant outcomes. We propose the HSVS model for grouped variable selection, where simultaneous selection of both groups and within-group variables is of interest. The HSVS model utilizes a discrete mixture prior distribution for group selection and group-specific Bayesian lasso hierarchies for variable selection within groups. We further provide methods for accounting for serial correlations within groups that incorporate Bayesian fused lasso methods for within-group selection. Finally, we propose a Bayesian method of estimating high-dimensional covariance matrices that can be decomposed into a low rank and sparse component. This covariance structure has a wide range of applications including factor analytical model and random effects model. We model the covariance matrices with the decomposition structure by representing the covariance model in the form of a factor analytic model where the number of latent factors is unknown. We introduce binary indicators for estimating the rank of the low rank component combined with a Bayesian graphical lasso method for estimating the sparse component. We further extend our method to a graphical factor analytic model where the graphical model of the residuals is of interest. We achieve sparse estimation of the inverse covariance of the residuals in the graphical factor model by employing a hyper-inverse Wishart prior method for a decomposable graph and a Bayesian graphical lasso method for an unrestricted graph.

The main goal of this thesis study is the optimization of the basic design parameters of hybrid electric vehicle drivetrain components to minimize fuel consumption and emission objectives, together with constraints derived from performance requirements. In order to generate a user friendly and flexible platform to model, select drivetrain components, simulate performance, and optimize parameters of series and parallel hybrid electric vehicles, a MATLAB based graphical user interface is designed. A basic sizing procedure for the internal combustion engine, electric motor, and battery is developed. Pre-defined control strategies are implemented for both types of hybrid configurations. To achieve better fuel consumption and emission values, while satisfying nonlinear performance constraints, multi-objective gradient based optimization procedure is carried out with user defined upper and lower bounds of optimization parameters. The optimization process is applied to a number of case studies and the results are evaluated by comparison with similar cases found in literature.

Computational vision and image analysis is a multidisciplinary scientific field that aims to make computers "see" in a way that is comparable to human perception. It is currently one of the most challenging research areas in artificial intelligence. In this regard, the extraction of information from the vast amount of visual data that are available today as well as the exploitation of the resulting information space becomes one of the greatest challenges in our days. To address such a challenge, this thesis describes a very general computational framework that can be used for performing efficient inference and learning for visual perception based on very rich and powerful models.

The image analysis company Izolde was in need of a user friendly graphical user interface (GUI) to use as a modifiable template to be able to meet a variety of requests and demands from their clientele. This paper describes the process of designing such a GUI with respect to theories within human computer interaction and available usability principles and theories.To familiarise and learn about other software on the market a background research was conducted. Wireframes as well as prototypes were created. With the help of recognised usability inspection tools tests were conducted on users with varied degree of computer experience. Test results were the basis for what would be altered and improved in terms of usability on the prototypes. The final result is a flexible user friendly GUI in regards to the criteria outlined by Izolde. / Detta examensarbete beskriver tillvägagångssättet i skapandet av en mall till ett användargränssnitt. Ett användargränssnitt är utseendet på ett program som användaren ser och interagerar med på en datorskärm, skärmen på en telefon eller annan typ av skärm. Användargränssnittet är utseendet av en programvara och det användaren visuellt kan tolka. I klartext tillåter ett användargränssnitt en användare att interagera med hårdvara genom inmatning (användarens påverkan på systemet) och utdata (resultat av användarens påverkan). Grafiska användargränssnitt skiljer sig från de textbaserade avändargränssnitten som tidigare var dominerande. Textbaserade lösningar känns ofta igen i äldre DOS-program och typiskt hos dessa är textinmatning av diverse kommandon och parametrar som leder till påverkan av systemet. Numer är det de grafiska användargränssnitten som dominerar. Fördelen med dessa är att de bygger på igenkänningsmekanismen hos människans tänkande och att de därför tillåter ett mer intuitivt användande av systemet i och med att de är just grafiska. Målet med examensarbetet var att designa en gränssnittsmall åt bildanalysföretaget Izolde. Gränssnittsmallen ska kunna modifieras för att kunna möta Izoldes kunders efterfrågan utifrån den typ av analys de vill kunna genomföra. För att kunna genomföra designprocessen krävdes djupare kunskap i vad användarvänlighet innebär och hur denna skall tillämpas i skapandet av ett grafiskt användargränssnitt. Med hjälp av vedertagna principer och teorier inom användarvänlighet var det möjligt att anta rollen som interaktionsdesigner. Interaktiondesign är en disciplin som definieras som beteendet och interaktionen mellan ett föremål, i detta fall ett grafiskt användargränssnitt, och dess användare. För att säkerställa att designprocessen skulle resultera i ett attraktivt och användarvänligt grafiskt användargränssnitt krävdes även att tester utfördes på användare med olika grad av datorkunskap på de wireframes och prototyper som skapades. En wireframe är en tidig skiss av designen på det slutgiltiga grafiska gränssnittet och kan beskrivas som en enkel ritning på ett papper eller en grov skiss skapat med lämplig programvara av vad designers vision av det slutgiltiga resultatet är. Genom att rita skisser skapar designern sig en god uppfattning om hur det grafiska gränssnittet kommer att presenteras i dess slutgiltiga skick. Dessutom är det mycket enklare och snabbare att göra ändringar på en skiss än i ett avancerat program som kräver mer exakt precision. Prototyper är mer sofistikerade versioner av wireframes och skapas i ett senare skede av designprocessen. Prototyper bär också större likheter med den slutliga produkten än wireframes. Designprocessen resulterade i ett flexibelt användarvänligt grafiskt användargränssnitt vars karaktär är anpassat i avseende på de kriterier som angetts av Izolde. Förhoppningen är att det skapade gränssnittet kommer kunna användas som en förändringsbar mall till ett gränssnitt som kan anpassas efter de kriterier och önskemål som ges av Izoldes kunder.

A set top box is an embedded device, much like a computer with limited capabilities. Its main purpose is to decode a video signal and output it to a TV. The set top box market is constantly growing and to be competitive in it, a set top box has to be able to do more than only TV. One way to make an attractive product is to give it an appealing user interface. This thesis is a part of a larger work at the company to find new ways to create graphical user interfaces. Its goal is to investigate what SVG implementations that exits, which one that is most suitable for an integration attempt and then perform the integration. Several SVG engines were investigated and one provided by the company was selected for integration. Three ways to integrate the SVG engine were identified. One of these alternatives was to extend the callback interface be- tween the engine and the underlying platform. Because of the good fit with the current architecture this alternative was chosen and implemented. As a part of this investigation a demo application suite of SVG content was also constructed. This investigation resulted in a working integration of the chosen SVG engine on the platform. It has also showed that SVG is a suitable language to build graphical user interfaces on set top boxes.

<p>This thesis is about Bayesian approaches for handling multiplicity. It considers three main kinds of multiple-testing scenarios: tests of exchangeable experimental units, tests for variable inclusion in linear regresson models, and tests for conditional independence in jointly normal vectors. Multiplicity adjustment in these three areas will be seen to have many common structural features. Though the modeling approach throughout is Bayesian, frequentist reasoning regarding error rates will often be employed.</p><p>Chapter 1 frames the issues in the context of historical debates about Bayesian multiplicity adjustment. Chapter 2 confronts the problem of large-scale screening of functional data, where control over Type-I error rates is a crucial issue. Chapter 3 develops new theory for comparing Bayes and empirical-Bayes approaches for multiplicity correction in regression variable selection. Chapters 4 and 5 describe new theoretical and computational tools for Gaussian graphical-model selection, where multiplicity arises in performing many simultaneous tests of pairwise conditional independence. Chapter 6 introduces a new approach to sparse-signal modeling based upon local shrinkage rules. Here the focus is not on multiplicity per se, but rather on using ideas from Bayesian multiple-testing models to motivate a new class of multivariate scale-mixture priors. Finally, Chapter 7 describes some directions for future study, many of which are the subjects of my current research agenda.</p> / Dissertation