Abstract
A key characteristic of simultaneous fault diagnosis is that the features extracted from the original patterns are strongly dependent. This paper proposes a new model of Bayesian classifier, which removes the fundamental assumption of naive Bayesian, i.e., the independence among features. In our model, the optimal bandwidth selection is applied to estimate the class-conditional probability density function (p.d.f.), which is the essential part of joint p.d.f. estimation. Three well-known indices, i.e., classification accuracy, area under ROC curve, and probability mean square error, are used to measure the performance of our model in simultaneous fault diagnosis. Simulations show that our model is significantly superior to the traditional ones when the dependence exists among features. © 2013 Elsevier Inc. All rights reserved.
| Original language | English |
|---|---|
| Pages (from-to) | 252-268 |
| Journal | Information Sciences |
| Volume | 259 |
| Early online date | 13 Sept 2013 |
| DOIs | |
| Publication status | Published - 20 Feb 2014 |
| Externally published | Yes |
Bibliographical note
The authors thank the editors and anonymous reviewers. Their valuable and constructive comments and suggestions helped them in significantly improving this paper. The authors also thank Prof. James Liu for his instructions on the improvement of language quality.Funding
This research is supported by the National Natural Science Foundation of China (71371063, 61170040 and 60903089), by the Natural Science Foundation of Hebei Province (F2013201110, F2012201023 and F2011201063), by the Key Scientific Research Foundation of Education Department of Hebei Province (ZD2010139), and by the CRG grant G-YL14 of The Hong Kong Polytechnic University.
Keywords
- Bayesian classification
- Dependent feature
- Joint probability density estimation
- Optimal bandwidth
- Simultaneous fault diagnosis
- Single fault