 |
Improved Ensemble Training for Hidden Markov Models using Random Relative Node Permutations
Davis, Richard I. A. and Lovell, Brian C. (2003). Improved Ensemble Training for Hidden Markov Models using Random Relative Node Permutations. In: Lovell, Brian C. and Maeder, Anthony J., Proceedings of the 2003 APRS Workshop on Digital Image Computing. The 2003 APRS Workshop on Digital Image Computing, Brisbane, (83-86). 7 February, 2003.
|
|
| |
| Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials) |
| Name |
Description |
MIMEType |
Size |
Downloads |
n83.pdf
|
n83.pdf |
application/pdf |
70.13KB |
243 |
| Author(s) |
Davis, Richard I. A.
Lovell, Brian C.
|
| Title of paper |
Improved Ensemble Training for Hidden Markov Models using Random Relative Node Permutations
|
| Conference name |
The 2003 APRS Workshop on Digital Image Computing
|
| Conference location |
Brisbane
|
| Conference dates |
7 February, 2003
|
| Proceedings title |
Proceedings of the 2003 APRS Workshop on Digital Image Computing
|
| Editor(s) |
Lovell, Brian C.
Maeder, Anthony J.
|
| Place published |
Brisbane
|
| Publisher |
Australian Pattern Recognition Society
|
| Publication date |
2003
|
| Volume number |
1
|
| Issue number |
1
|
| ISBN |
0-9580255-2-5
|
| Start page |
83
|
| End page |
86
|
| Language |
eng
|
| Abstract/Summary |
Hidden Markov Models have many applications in signal processing and pattern recognition, but their convergence-based training algorithms are known to suffer from oversensitivity to the initial random model choice. This paper focuses upon the use of model averaging, ensemble thresholding, and random relative model permutations for improving average model performance. A method is described which trains by searching for the best relative permutation set for ensemble averaging. This uses the fit to the training set as an indicator. The work provides a simpler alternative to previous permutation-based ensemble averaging methods.
|
| Subjects |
280207 Pattern Recognition
|
| Keyword(s) |
iris-research
Markov training
learning
Hidden Markov models
|
|
|
|
