Modified Baum-Welch Algorithm Research Articles

An application of a semi-hidden Markov model in wireless communication systems

Stochastic processes are approved presentation of real systems which its development in space or time can be supposed as random. A semi-hidden Markov model as a type of stochastic processes is a modification of hidden Markov models with states that are no longer totally unobservable and are less hidden. This mathematical model is employed for modeling data sequences with long runs, memory and statistical inertia. In this article, we investigate the theory of the semi-hidden Markov model along with its parameter estimation and order estimation methods. Moreover, the proposed model is applied to model the error traces generated by the wireless channels. A new Markov-based trace analysis algorithm is suggested to divide a non-stationary network error trace into stationary parts. By means of the best semi-hidden Markov model and fitting probability distribution, we would be able to model these parts accurately. Calculating the information measure criteria and the autocorrelation function by running the modified Baum–Welch algorithm several times help us to find the optimal order of the semi-hidden Markov model.

ON A LOGISTIC-HIDDEN MARKOV MODEL FOR COMMUNICATION SYSTEMS

This article proposes an extension of hidden Markov models for communication systems by allowing the Markovian transitions between the channel's states to be influenced by an external “catalyzer” (i.e., environmental or experimental conditions). The stochastic influence of the catalyzer is expressed by logistic link functions. The model can be useful for modeling communication channels, for example, in predicting single-ion channel behavior for cell signaling. A simulation study is provided in order to explore how trainable the model is. We prove that the Logistic-Hidden Markov model can be very well trained by means of the maximum likelihood method, assisted by a modified Baum–Welch algorithm.

System and method for training a class-specific hidden markov model using a modified baum-welch algorithm

A system and method for training a class-specific hidden Markov model (HMM) is used for modeling physical phenomena, such as speech, characterized by a finite number of states. The method receives training data and estimates parameters of the class-specific HMM from the training data using a modified Baum-Welch algorithm, which uses likelihood ratios with respect to a common state (e.g., noise) and based on sufficient statistics for each state. The parameters are stored for use in processing signals representing the physical phenomena, for example, in speech processing applications. The modified Baum-Welch algorithm is an iterative algorithm including class-specific forward and backward procedures and HMM reestimation formulas.

A constrained HMM-based approach to the estimation of perceptual switching dynamics in pigeons

We describe a method for the analysis of behavioural data from experiments on perception of bistable stimuli in pigeons. The approach is based on a Hidden Markov Model (HMM) with additional linear factorial constraints for which a modified Baum–Welch algorithm is derived. It allows the estimation of the perceptual switching events, which might directly relate to transitions between states of activation of corresponding neural populations. From the resulting time series, characteristics of the underlying perceptual dynamics can be estimated. We also demonstrate that—inspite of the Markov assumption—the method can reveal certain non-Markovian contributions to the dynamics.

A modified Baum-Welch algorithm for hidden Markov models with multiple observation spaces

We derive an algorithm similar to the well-known Baum-Welch (1970) algorithm for estimating the parameters of a hidden Markov model (HMM). The new algorithm allows the observation PDF of each state to be defined and estimated using a different feature set. We show that estimating parameters in this manner is equivalent to maximizing the likelihood function for the standard parameterization of the HMM defined on the input data space. The processor becomes optimal if the state-dependent feature sets are sufficient statistics to distinguish each state individually from a common state.

Unconstrained Farsi handwritten word recognition using fuzzy vector quantization and hidden Markov models

An unconstrained Farsi handwritten word recognition system based on fuzzy vector quantization (FVQ) and hidden Markov model (HMM) for reading city names in postal addresses is presented. Preprocessing techniques including binarization, noise removal, slope correction and baseline estimation are described. Each word image is represented by its contour information. The histogram of chain code slopes of the image strips (frames), scanned from right to left by a sliding window, is used as feature vectors. Fuzzy c-means (FCM) clustering is used for generating a fuzzy codebook. A separate HMM is trained by modified Baum–Welch algorithm for each city name. A test image is recognized by finding the best match (likelihood) between the image and all of the HMM word models using forward algorithm. Experimental results show the advantages of using FVQ/HMM recognizer engine instead of conventional discrete HMMs.