Continuously variable duration hidden Markov models for automatic speech recognition
Computer Speech and Language
On hidden fractal model signal processing
Signal Processing
State duration modelling in hidden Markov models
Signal Processing
Nonstationary hidden Markov model
Signal Processing
Frontiers in queueing
An integrated mobility and traffic model for resource allocation in wireless networks
WOWMOM '00 Proceedings of the 3rd ACM international workshop on Wireless mobile multimedia
Estimation of nonstationary hidden Markov models by MCMC sampling
ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 03
PM2.5 concentration prediction using hidden semi-Markov model-based times series data mining
Expert Systems with Applications: An International Journal
The use of hidden semi-Markov models in clinical diagnosis maze tasks
Intelligent Data Analysis
Rigorous statistical analysis of internet loss measurements
Proceedings of the ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Text entry for mobile devices using ad-hoc abbreviation
Proceedings of the International Conference on Advanced Visual Interfaces
Optimizing two-dimensional search results presentation
Proceedings of the fourth ACM international conference on Web search and data mining
Unsupervised segmentation of hidden semi-Markov non-stationary chains
Signal Processing
HMM-based characterization of channel behavior for networked control systems
Proceedings of the 1st international conference on High Confidence Networked Systems
Information integration over time in unreliable and uncertain environments
Proceedings of the 21st international conference on World Wide Web
Towards "live" synthetic populations for large-scale realistic multiagent simulations
Proceedings of the 2013 international conference on Autonomous agents and multi-agent systems
Rigorous statistical analysis of internet loss measurements
IEEE/ACM Transactions on Networking (TON)
| Hi-index | 0.08 |
A hidden Markov model (HMM) encompasses a large class of stochastic process models and has been successfully applied to a number of scientific and engineering problems, including speech and other pattern recognition problems, and DNA sequence comparison. A hidden semi-Markov model (HSMM) is an extension of HMM, designed to remove the constant or geometric distributions of the state durations assumed in HMM. A larger class of practical problems can be appropriately modeled in the setting of HSMM. A major restriction is found, however, in both conventional HMM and HSMM, i.e., it is generally assumed that there exists at least one observation associated with every state that the hidden Markov chain takes on. We will remove this assumption and consider the following situations: (i) observation data may be missing for some intervals; and (ii) there are multiple observation streams that are not necessarily synchronous to each other and may have different "emission distributions" for the same state. We propose a new and computationally efficient forward-backward algorithm for HSMM with missing observations and multiple observation sequences. The required computational amount for the forward and backward variables is reduced to O(D), where D is the maximum allowed duration in a state. Finally, we will apply the extended HSMM to estimate the mobility model parameters for the Internet service provisioning in wireless networks.