Search

Your search keyword '"Liu, Weiyi"' showing total 72 results
Start Over Author "Liu, Weiyi" Publisher ieee
72 results on '"Liu, Weiyi"'

14. A method of backward probabilistic logic reasoning of decision problems

Author

Liu Weiyi and Li Yong

Subjects
Reasoning system, symbols.namesake, Theoretical computer science, Markov chain, Conditional event algebra, Probabilistic logic, symbols, Markov process, Discrete event simulation, Decision problem, Mathematics, Event (probability theory)
Abstract

There are critical problems in reasoning of decision problems with casual relations. One is how to process backward reasoning. Another is the discrepancy between logic and probability in denoting casual relations. We propose and implement a backward probabilistic logic reasoning approach, which combines conditional event algebra and Markov Monte Carlo simulating algorithm. By partly changing casual relations in a decision problem, we first make the problem of backward reasoning possible, and then bring the logic consistent with the probability in denoting casual relation by extending normal measurable space with conditional event. We transform the conditional event to normal events and corresponding logical combination events via conditional event algebra, and use Gibbs simulation to sample the normal events to be a stationary state. By computing the quantitative values of the stationary events, we can evaluate the quantitative value of conjunction and disjunction operations of conditional event at last and finish backward reasoning. An application of our method shows how we process a backward reasoning in a sequential decision.

Published
2008

19. A Parallel and Incremental Approach for Data-Intensive Learning of Bayesian Networks.

Author

Yue, Kun, Fang, Qiyu, Wang, Xiaoling, Li, Jin, and Liu, Weiyi

Abstract

Bayesian network (BN) has been adopted as the underlying model for representing and inferring uncertain knowledge. As the basis of realistic applications centered on probabilistic inferences, learning a BN from data is a critical subject of machine learning, artificial intelligence, and big data paradigms. Currently, it is necessary to extend the classical methods for learning BNs with respect to data-intensive computing or in cloud environments. In this paper, we propose a parallel and incremental approach for data-intensive learning of BNs from massive, distributed, and dynamically changing data by extending the classical scoring and search algorithm and using MapReduce. First, we adopt the minimum description length as the scoring metric and give the two-pass MapReduce-based algorithms for computing the required marginal probabilities and scoring the candidate graphical model from sample data. Then, we give the corresponding strategy for extending the classical hill-climbing algorithm to obtain the optimal structure, as well as that for storing a BN by pairs. Further, in view of the dynamic characteristics of the changing data, we give the concept of influence degree to measure the coincidence of the current BN with new data, and then propose the corresponding two-pass MapReduce-based algorithms for BNs incremental learning. Experimental results show the efficiency, scalability, and effectiveness of our methods. [ABSTRACT FROM PUBLISHER]

Published
2015
Full Text
View/download PDF

48. Markov Chain Approximation Algorithm for Event-Based State Estimation.

Author

Lee, Sangjin, Liu, Weiyi, and Hwang, Inseok

Subjects
MARKOV processes, APPROXIMATION algorithms, NONLINEAR control theory, STOCHASTIC differential equations, ALGORITHMS, UNCERTAIN systems
Abstract

This brief presents a general framework for the continuous-time nonlinear event-based state estimation problem. Using the information from observations made by event-based sampling, the goal of the event-based estimation problem is to estimate the state of stochastic differential equations which represent the uncertain system dynamics. This problem is challenging because measurements are taken only if some events happen rather than with a fixed sampling interval. In this brief, a theoretical solution for the event-based state estimation problem is derived and a numerical algorithm based on Markov chain approximation is proposed. The proposed algorithm for the event-based state estimation is demonstrated with an illustrative example. [ABSTRACT FROM PUBLISHER]

Published
2015
Full Text
View/download PDF

50. A stochastic approximation based state estimation algorithm for Stochastic Hybrid Systems.

Author

Liu, Weiyi and Hwang, Inseok

Abstract

This paper is focused on the state estimation for the Stochastic Hybrid System (SHS) which is a class of continuous-time stochastic processes with the interacting continuous and discrete dynamics. The state estimation problem considered in this paper involves computing the probability distributions of both the continuous and the discrete state of a SHS with the information given by a series of noisy discrete-time observations from sensors at each sampling time. The numerical state estimation algorithm proposed in this paper is based on a stochastic approximation approach by using a Markov Chain (MC) to approximate the dynamics of the SHS and thus estimates the state of the MC instead of the SHS. The proposed algorithm is validated through a scenario of aircraft tracking for air traffic control. [ABSTRACT FROM PUBLISHER]

Published
2012
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results