Results 1  10
of
16
The Bayes Net Toolbox for MATLAB
 Computing Science and Statistics
, 2001
"... The Bayes Net Toolbox (BNT) is an opensource Matlab package for directed graphical models. BNT supports many kinds of nodes (probability distributions), exact and approximate inference, parameter and structure learning, and static and dynamic models. BNT is widely used in teaching and research: the ..."
Abstract

Cited by 216 (1 self)
 Add to MetaCart
The Bayes Net Toolbox (BNT) is an opensource Matlab package for directed graphical models. BNT supports many kinds of nodes (probability distributions), exact and approximate inference, parameter and structure learning, and static and dynamic models. BNT is widely used in teaching and research: the web page has received over 28,000 hits since May 2000. In this paper, we discuss a broad spectrum of issues related to graphical models (directed and undirected), and describe, at a highlevel, how BNT was designed to cope with them all. We also compare BNT to other software packages for graphical models, and to the nascent OpenBayes effort.
Learning and understanding dynamic scene activity: a review
, 2003
"... We are entering an era of more intelligent cognitive vision systems. Such systems can analyse activity in dynamic scenes to compute conceptual descriptions from motion trajectories of moving people and the objects they interact with. Here we review progress in the development of flexible, generative ..."
Abstract

Cited by 61 (0 self)
 Add to MetaCart
We are entering an era of more intelligent cognitive vision systems. Such systems can analyse activity in dynamic scenes to compute conceptual descriptions from motion trajectories of moving people and the objects they interact with. Here we review progress in the development of flexible, generative models that can explain visual input as a combination of hidden variables and can adapt to new types of input. Such models are particularly appropriate for the tasks posed by cognitive vision as they incorporate learning as well as having sufficient structure to represent a general class of problems. In addition, generative models explain all aspects of the input rather than attempting to ignore irrelevant sources of variation as in exemplarbased learning. Applications of these models in visual interaction for education, smart rooms and cars, as well as surveillance systems is also briefly reviewed.
Generative Models for Learning and Understanding Dynamic Scene Activity
 in ECCV Workshop on Generative Model Based Vision
, 2002
"... We are entering an era of more intelligent cognitive vision systems. Such systems can analyse activity in dynamic scenes to compute conceptual descriptions from motion trajectories of moving people and the objects they interact with. Here we review progress in the development of flexible, generative ..."
Abstract

Cited by 23 (2 self)
 Add to MetaCart
(Show Context)
We are entering an era of more intelligent cognitive vision systems. Such systems can analyse activity in dynamic scenes to compute conceptual descriptions from motion trajectories of moving people and the objects they interact with. Here we review progress in the development of flexible, generative models that can explain visual input as a combination of hidden variables and can adapt to new types of input. Such models are particularly appropriate for the tasks posed by cognitive vision as they incorporate learning as well as having sufficient structure to represent a general class of problems. In addition, generative models explain all aspects of the input rather than attempting to ignore irrelevant sources of variation as in exemplarbased learning. Applications of these models in visual interaction for education, smart rooms and cars, as well as surveillance systems is also briefly reviewed.
Structured priors for structure learning
 In Proceedings of the 22nd Conference on Uncertainty in Artificial Intelligence (UAI
, 2006
"... Traditional approaches to Bayes net structure learning typically assume little regularity in graph structure other than sparseness. However, in many cases, we expect more systematicity: variables in realworld systems often group into classes that predict the kinds of probabilistic dependencies they ..."
Abstract

Cited by 21 (9 self)
 Add to MetaCart
(Show Context)
Traditional approaches to Bayes net structure learning typically assume little regularity in graph structure other than sparseness. However, in many cases, we expect more systematicity: variables in realworld systems often group into classes that predict the kinds of probabilistic dependencies they participate in. Here we capture this form of prior knowledge in a hierarchical Bayesian framework, and exploit it to enable structure learning and type discovery from small datasets. Specifically, we present a nonparametric generative model for directed acyclic graphs as a prior for Bayes net structure learning. Our model assumes that variables come in one or more classes and that the prior probability of an edge existing between two variables is a function only of their classes. We derive an MCMC algorithm for simultaneous inference of the number of classes, the class assignments of variables, and the Bayes net structure over variables. For several realistic, sparse datasets, we show that the bias towards systematicity of connections provided by our model can yield more accurate learned networks than the traditional approach of using a uniform prior, and that the classes found by our model are appropriate. 1
UNDERSTANDING PATHWAYS
"... The challenge with todays microarray experiments is to infer biological conclusions from them. There are two crucial difficulties to be surmounted in this challenge:(1) A lack of suitable biological repository that can be easily integrated into computational algorithms. (2) Contemporary algorithms u ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
The challenge with todays microarray experiments is to infer biological conclusions from them. There are two crucial difficulties to be surmounted in this challenge:(1) A lack of suitable biological repository that can be easily integrated into computational algorithms. (2) Contemporary algorithms used to analyze microarray data are unable to draw consistent biological results from diverse datasets of the same disease. To deal with the first difficulty, we believe a core database that unifies available biological repositories is important. Towards this end, we create a unified biological database from three popular biological repositories (KEGG, Ingenuity and Wikipathways). This database provides computer scientists the flexibility of easily integrating biological information using simple API calls or SQL queries. To deal with the second difficulty of deriving consistent biological results from the experiments, we first conceptualize the notion of “subnetworks”, which refers to a
Bayesian Belief Networks for Dementia Diagnosis and Other Applications: A Comparison of HandCrafting and Construction using A Novel Data Driven Technique
, 2008
"... The Bayesian network (BN) formalism is a powerful representation for encoding domains characterised by uncertainty. However, before it can be used it must first be constructed, which is a major challenge for any reallife problem. There are two broad approaches, namely the handcrafted approach, whi ..."
Abstract
 Add to MetaCart
(Show Context)
The Bayesian network (BN) formalism is a powerful representation for encoding domains characterised by uncertainty. However, before it can be used it must first be constructed, which is a major challenge for any reallife problem. There are two broad approaches, namely the handcrafted approach, which relies on a human expert, and the datadriven approach, which relies on data. The former approach is useful, however issues such as human bias can introduce errors into the model. We have conducted a literature review of the expertdriven approach, and we have cherrypicked a number of common methods, and engineered a framework to assist nonBN experts with expertdriven construction of BNs. The latter construction approach uses algorithms to construct the model from a data set. However, construction from data is provably NPhard [45]. To solve this problem, approximate, heuristic algorithms have been proposed; in particular, algorithms that assume an order between the nodes, therefore reducing the search space. However, traditionally, this approach relies on an expert providing the order
Bayesian Network Classifier for Medical Data Analysis
"... Abstract: Bayesian networks encode causal relations between variables using probability and graph theory. They can be used both for prediction of an outcome and interpretation of predictions based on the encoded causal relations. In this paper we analyse a treelike Bayesian network learning algorit ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract: Bayesian networks encode causal relations between variables using probability and graph theory. They can be used both for prediction of an outcome and interpretation of predictions based on the encoded causal relations. In this paper we analyse a treelike Bayesian network learning algorithm optimised for classification of data and we give solutions to the interpretation and analysis of predictions. The classification of logical – i.e. binary – data arises specifically in the field of medical diagnosis, where we have to predict the survival chance based on different types of medical observations or we must select the most relevant cause corresponding again to a given patient record. Surgery survival prediction was examined with the algorithm. Bypass surgery survival chance must be computed for a given patient, having a dataset of 66 medical examinations for 313 patients.