Fast approximation of Kullback-Leibler distance for dependence trees and hidden Markov models
Author:
- 20 March 2003
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 10 (4) , 115-118
Abstract
We present a fast algorithm to approximate the Kullback-Leibler distance (KLD) between two dependence tree models. The algorithm uses the "upward" (or "forward") procedure to compute an upper bound for the KLD. For hidden Markov models, this algorithm is reduced to a simple expression. Numerical experiments show that for a similar accuracy, the proposed algorithm offers a saving of hundreds of times in computational complexity compared to the commonly used Monte Carlo method. This makes the proposed algorithm important for real-time applications, such as image retrieval.Keywords
This publication has 6 references indexed in Scilit:
- Rotation invariant texture characterization and retrieval using steerable wavelet-domain hidden Markov modelsIEEE Transactions on Multimedia, 2002
- Elements of Information Theory2001
- Wavelet-based statistical signal processing using hidden Markov modelsIEEE Transactions on Signal Processing, 1998
- Parameter estimation of dependence tree models using the EM algorithmIEEE Signal Processing Letters, 1995
- A tutorial on hidden Markov models and selected applications in speech recognitionProceedings of the IEEE, 1989
- A Probabilistic Distance Measure for Hidden Markov ModelsAT&T Technical Journal, 1985