Results 11  20
of
32
Online Probability Density Estimation of Nonstationary Random Signal using Dynamic Bayesian Networks
"... Abstract: We present two estimators for discrete nonGaussian and nonstationary probability density estimation based on a dynamic Bayesian network (DBN). The first estimator is for offline computation and consists of a DBN whose transition distribution is represented in terms of kernel functions. T ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Abstract: We present two estimators for discrete nonGaussian and nonstationary probability density estimation based on a dynamic Bayesian network (DBN). The first estimator is for offline computation and consists of a DBN whose transition distribution is represented in terms of kernel functions. The estimator parameters are the weights and shifts of the kernel functions. The parameters are determined through a recursive learning algorithm using maximum likelihood (ML) estimation. The second estimator is a DBN whose parameters form the transition probabilities. We use an asymptotically convergent, recursive, online algorithm to update the parameters using observation data. The DBN calculates the state probabilities using the estimated parameters. We provide examples that demonstrate the usefulness and simplicity of the two proposed estimators.
Prognosis of HighGrade Carcinoid Tumor Patients using Dynamic LimitedMemory Influence Diagrams ∗
"... Dynamic limitedmemory influence diagrams (DLIMIDs) have been developed as a framework for decisionmaking under uncertainty over time. We show that DLIMIDs constructed from twostage temporal LIMIDs can represent infinitehorizon decision processes. Given a treatment strategy supplied by the physicia ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Dynamic limitedmemory influence diagrams (DLIMIDs) have been developed as a framework for decisionmaking under uncertainty over time. We show that DLIMIDs constructed from twostage temporal LIMIDs can represent infinitehorizon decision processes. Given a treatment strategy supplied by the physician, DLIMIDs may be used as prognostic models. The theory is applied to determine the prognosis of patients that suffer from an aggressive type of neuroendocrine tumor. 1
2.2 Bayes Theorem........................................ 4
"... This thesis describes an approach that uses reinforcement learning together with distributed perception network fusion systems in order to perform mobile sensor control. A case study of such mobile sensors is the chemical leak detection problem. The proposed solution deals with partial observability ..."
Abstract
 Add to MetaCart
(Show Context)
This thesis describes an approach that uses reinforcement learning together with distributed perception network fusion systems in order to perform mobile sensor control. A case study of such mobile sensors is the chemical leak detection problem. The proposed solution deals with partial observability of the true state and makes use of linear function approximation to learn a value function that maps beliefaction pairs into values, distributed perception networks to create a correct, robust and computationally efficient system for the inference of gas leak given sensor observations, and bayesian inference to estimate the leak location. Results show that by keeping a continuous belief represented through entropies and representing information about both single cell and aggregation of cells, the system can learn successfully; furthermore, studies about the influence that the distributed perception networks give to the whole system are done. Discussions and future
Hybrid SymbolicProbabilistic Plan Recognizer: Initial steps
"... It is important for agents to model other agents ’ unobserved plans and goals, based on their observable actions, a process known as plan recognition. Plan recognition often takes the form of matching observations of an agent’s actions to a planlibrary, a model of possible plans selected by the agen ..."
Abstract
 Add to MetaCart
It is important for agents to model other agents ’ unobserved plans and goals, based on their observable actions, a process known as plan recognition. Plan recognition often takes the form of matching observations of an agent’s actions to a planlibrary, a model of possible plans selected by the agent. In this paper, we present efficient algorithms that handle a number of key capabilities implied by plan recognition applications, in the context of hybrid symbolicprobabilistic recognizer. The central idea behind the hybrid approach is to combine the symbolic approach with probabilistic inference: the symbolic recognizer efficiently filters inconsistent hypotheses, passing only the consistent hypotheses to a probabilistic inference engine. There are few investigations that utilize an hybrid symbolicprobabilistic approach. The advantage of this kind of inference is potentially enormous. First, it can be highly efficient. Second, it can efficiently deal with richer class of plan recognition challenges, such as recognition based on duration of behaviors, recognition despite intermittently lost observations, and recognition of interleaved plans.
Unsupervised Models for Spatial, Temporal and Relational Systems
, 2009
"... Social processes can be strongly influenced by their spatial and temporal environment, as well as relational structures specific to the process itself. While it has traditionally been expedient to study one or two of these dimensions at a time, it is increasingly feasible to collect data necessary t ..."
Abstract
 Add to MetaCart
(Show Context)
Social processes can be strongly influenced by their spatial and temporal environment, as well as relational structures specific to the process itself. While it has traditionally been expedient to study one or two of these dimensions at a time, it is increasingly feasible to collect data necessary to investigate how, and in what combinations and proportions spatial, temporal and relational (STR) factors govern a process. This proposal is concerned with enabling the early stages of such an analysis, in which the researcher has a hypothesis regarding what relationships exist between STR variables, but not the details and relative strengths of these relationships. Can we express this generalized hypothesis, and algorithmically use available data to recommend a more specific one? I adopt probabilistic graphical models (PGMs) as a flexible framework for representing structural hypotheses, and introduce a templating system for generating regular PGM structures appropriate STR data. In fitting these models to data, I argue against both supervised training and Bayesian unsupervised methods, suggesting a focus on fast, useful inference over (even approximate) optimality. To this end, I introduce Expectation Maximizing belief propagation (EMBP) algorithms, which perform fast unsupervised learning in graphical models with spatial, temporal and relational structure, leading to a variety of
Classification of Abnormal Activities in Video
"... In multimedia computing the recognition of abnormal activities is becoming a major area of research interest. With applications in humancomputerinteraction, elder care, security, and surveillance there is a strong push for advances in our ability to recognize both normal and abnormal activities at ..."
Abstract
 Add to MetaCart
(Show Context)
In multimedia computing the recognition of abnormal activities is becoming a major area of research interest. With applications in humancomputerinteraction, elder care, security, and surveillance there is a strong push for advances in our ability to recognize both normal and abnormal activities at the semantic level. We use a probabilistic, hierarchical representation of activities to do recognition and provide an automatic way to define the lowlevel states. We classify abnormal activities meaningfully in terms of known highlevel activities and show brief results of this work.
DIALOG ACT TAGGING USING GRAPHICAL MODELS
"... Detecting discourse patterns such as dialog acts (DAs) is an important factor for processing spoken conversations and meetings. Different techniques have been used to tag dialog acts in the past such as hidden Markov models and neural networks. In this work, a full analysis of dialog act tagging usi ..."
Abstract
 Add to MetaCart
(Show Context)
Detecting discourse patterns such as dialog acts (DAs) is an important factor for processing spoken conversations and meetings. Different techniques have been used to tag dialog acts in the past such as hidden Markov models and neural networks. In this work, a full analysis of dialog act tagging using different generative and conditional dynamic Bayesian networks (DBNs) is performed, where both conventional switching ngrams and factored language models (FLMs) are used as DBN edge implementations. Our tests on the ICSI meeting recorder dialog act (MRDA) corpus show that the factored language model implementations are better than the switching ngram approach. Our results also show that by using virtual evidence, the label bias problem in conditional models can be avoided. Also, we find that on a corpus such as MRDA, using the dialog acts of previous sentences to help predict current words does not improve our conditional model. 1.
Abstract
"... Stability is a desirable characteristic for linear dynamical systems, but it is often ignored by algorithms that learn these systems from data. We propose a novel method for learning stable linear dynamical systems: we formulate an approximation of the problem as a convex program, start with a solut ..."
Abstract
 Add to MetaCart
(Show Context)
Stability is a desirable characteristic for linear dynamical systems, but it is often ignored by algorithms that learn these systems from data. We propose a novel method for learning stable linear dynamical systems: we formulate an approximation of the problem as a convex program, start with a solution to a relaxed version of the program, and incrementally add constraints to improve stability. Rather than continuing to generate constraints until we reach a feasible solution, we test stability at each step; because the convex program is only an approximation of the desired problem, this early stopping rule can yield a higherquality solution. We apply our algorithm to the task of learning dynamic textures from image sequences as well as to modeling biosurveillance drugsales data. The constraint generation approach leads to noticeable improvement in the quality of simulated sequences. We compare our method to those of Lacy and Bernstein [1, 2], with positive results in terms of accuracy, quality of simulated sequences, and efficiency. 1
A STOCHASTIC APPROACH FOR CREATING DYNAMIC CONTEXTAWARE SERVICES IN SMART HOME ENVIRONMENT
"... Contextaware service platforms aim at providing users with dynamic services that adapt to changeful environment. Though various toolkits have been developed, it is still not easy for endusers to program their own personalized services. In this paper, we present a contextaware service platform call ..."
Abstract
 Add to MetaCart
(Show Context)
Contextaware service platforms aim at providing users with dynamic services that adapt to changeful environment. Though various toolkits have been developed, it is still not easy for endusers to program their own personalized services. In this paper, we present a contextaware service platform called Synapse that can adequately relate users ’ contexts and services, and automatically generate contextaware services based on users ’ habits. By exploiting the recorded histories of contexts and services, Synapse can learn different users ’ habits. Then Synapse can predict the most relevant services that users will use in current situation based on their habits, and provide services in different modes. As learning and predicting mechanisms are on the basis of a stochastic approach – Bayesian Networks [5], Synapse can absorb various uncertainties arising from sensor data and provide truly personalized services. 1.