Sciweavers

410 search results - page 25 / 82
» Incremental Construction of Structured Hidden Markov Models
Sort
View
ESWA
2007
117views more  ESWA 2007»
13 years 10 months ago
An HMM for detecting spam mail
Hidden Markov Models, or HMMs for short, have been recently used in Bioinformatics for the classification of DNA or protein chains, giving rise to what is known as Profile Hidde...
José Gordillo, Eduardo Conde
PR
2010
147views more  PR 2010»
13 years 9 months ago
Minimum classification error learning for sequential data in the wavelet domain
Wavelet analysis has found widespread use in signal processing and many classification tasks. Nevertheless, its use in dynamic pattern recognition have been much more restricted ...
D. Tomassi, Diego H. Milone, L. Forzani
ECAI
2004
Springer
14 years 4 months ago
Learning Complex and Sparse Events in Long Sequences
The Hierarchical Hidden Markov Model (HHMM) is a well formalized tool suitable to model complex patterns in long temporal or spatial sequences. Even if effective algorithms are ava...
Marco Botta, Ugo Galassi, Attilio Giordana
ICML
2009
IEEE
14 years 11 months ago
Exploiting sparse Markov and covariance structure in multiresolution models
We consider Gaussian multiresolution (MR) models in which coarser, hidden variables serve to capture statistical dependencies among the finest scale variables. Tree-structured MR ...
Myung Jin Choi, Venkat Chandrasekaran, Alan S. Wil...
ICPR
2008
IEEE
14 years 12 months ago
Automatic generation of HMM topology for sign language recognition
Sign language is used for communicating to people with hearing difficulties. Recogntion of a sign language image sequence is challenging because of the variety of hand shapes and ...
Tadashi Matsuo, Yoshiaki Shirai, Nobutaka Shimada