On-line probabilistic classification with particle filters

Author

Højen-Sorensen, Pedro A d F R ; De Freitas, Nando ; Fog, Torben

Author_Institution

Dept. of Math. Modelling, Tech. Univ. Denmark, Lyngby, Denmark

Volume

1

fYear

2000

fDate

2000

Firstpage

386

Abstract

We apply particle filters to the problem of on-line classification with possibly overlapping classes. This allows us to compute the probabilities of class membership as the classes evolve. Although we adopt neural network classifiers, the work can be extended to any other parametric classification scheme. We demonstrate our methodology on a simple example and on the problem of fault detection of dynamically operated marine diesel engines

Keywords

fault location; filtering theory; internal combustion engines; multilayer perceptrons; pattern classification; class membership; condition monitoring; dynamically operated marine diesel engines; fault detection; multi-layer perceptrons; neural network classifiers; online probabilistic classification; overlapping classes; parametric classification; particle filters; probabilities; real-time decision systems; sequential classification; Classification tree analysis; Diesel engines; Fault detection; Filtering; Gaussian noise; Hidden Markov models; Logistics; Neural networks; Particle filters; State-space methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks for Signal Processing X, 2000. Proceedings of the 2000 IEEE Signal Processing Society Workshop

Conference_Location

Sydney, NSW

ISSN

1089-3555

Print_ISBN

0-7803-6278-0

Type

conf

DOI

10.1109/NNSP.2000.889430

Filename

889430