Speaker:

Xin Wang, Department of Information Science

Title:

"Time-line" Hidden Markov Experts for Time Series Prediction

Location:

Archway 2

Abstract:

A modularised model, based on the Mixture of Experts (ME) algorithm for time series prediction, will be introduced in the talk. A set of connectionist models learn to be local experts over some phases of the trajectory in the state space, so the trajectory-regression task is divided into some simple ones and precise approximation is expected.

The dynamics for combining the local experts is a Hidden Markov Model (HMM), where the phases of the trajectory are regarded as the states of a HMM and the state transition on the Markov chain is the process of the underlying system evolving from one phase to another on the trajectory.

The transition probabilities of the hidden Markov model (HMM) are designed to be time-varying and a modified Baum-Welch algorithm is introduced for the learning of the "time-line" HMM.