Wang, Haiqin
(2007)
Building Bayesian Networks: Elicitation, Evaluation, and Learning.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
As a compact graphical framework for representation of multivariate probabilitydistributions, Bayesian networks are widely used for efficient reasoning underuncertainty in a variety of applications, from medical diagnosis to computertroubleshooting and airplane fault isolation. However, construction of Bayesiannetworks is often considered the main difficulty when applying this frameworkto real-world problems. In real world domains, Bayesian networks are often built by knowledge engineering approach. Unfortunately, eliciting knowledge from domain experts isa very time-consuming process, and could result in poor-quality graphicalmodels when not performed carefully. Over the last decade, the research focusis shifting more towards learning Bayesian networks from data, especially withincreasing volumes of data available in various applications, such asbiomedical, internet, and e-business, among others.Aiming at solving the bottle-neck problem of building Bayesian network models, thisresearch work focuses on elicitation, evaluation and learning Bayesiannetworks. Specifically, the contribution of this dissertation involves the research in the following five areas:a) graphical user interface tools forefficient elicitation and navigation of probability distributions, b) systematic and objective evaluation of elicitation schemes for probabilistic models, c)valid evaluation of performance robustness, i.e., sensitivity, of Bayesian networks,d) the sensitivity inequivalent characteristic of Markov equivalent networks, and the appropriateness of using sensitivity for model selection in learning Bayesian networks,e) selective refinement for learning probability parameters of Bayesian networks from limited data with availability of expert knowledge. In addition, an efficient algorithm for fast sensitivity analysis is developed based on relevance reasoning technique. The implemented algorithm runs very fast and makes d) and e) more affordable for real domain practice.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
15 October 2007 |
| Date Type: |
Completion |
| Defense Date: |
6 December 2004 |
| Approval Date: |
15 October 2007 |
| Submission Date: |
6 December 2004 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Dietrich School of Arts and Sciences > Intelligent Systems |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
Bayesian network; belief networks; elicitation; evaluation; learning; model building |
| Other ID: |
http://etd.library.pitt.edu/ETD/available/etd-12062004-114342/, etd-12062004-114342 |
| Date Deposited: |
10 Nov 2011 20:08 |
| Last Modified: |
15 Nov 2016 13:53 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/10084 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}