Two-timescale algorithms for simulation optimization of hidden Markov models
β Scribed by Shalabh Bhatnagar; Michael C. Fu; Steven I. Marcus; Shashank Bhatnagar
- Book ID
- 110252240
- Publisher
- Taylor and Francis Group
- Year
- 2001
- Tongue
- English
- Weight
- 207 KB
- Volume
- 33
- Category
- Article
- ISSN
- 0740-817X
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
For hidden Markov model (HMM) based speech recognition where the basic speech unit is smaller than the recognizer's output unit, the standard full Baum-Welch re-estimation procedure for the HMM training is very costly in computation. This is hecause it requires evaluation of the HMM output densities
## Background Secondary structure prediction is a useful first step toward 3D structure prediction. A number of successful secondary structure prediction methods use neural networks, but unfortunately, neural networks are not intuitively interpretable. On the contrary, hidden Markov models are grap