DEPARTMENT OF MEDICAL STATISTICS
DE MONTFORT UNIVERSITY, LEICESTER
Seminar Announcement
Time: Thursday, 25th February 1999, 1 p.m.
Place: James Went Building, DMU Leicester
Room JW 8-6/8-7
Speaker: Professor J.K. Lindsey
De Montfort University, Leicester
Title: An Introduction to Hidden Markov Models
Abstract
Hidden Markov models have come to be widely used in various areas such
as speech recognition and DNA sequence analysis. They belong to the
class of dynamic models that are updated over time using variants of
the Kalman filtering procedure. Typically, they have been used in
discrete time for discrete outcomes but can easily be extended, both
to continuous time and to any type of response outcome. These models
are especially appropriate when dependence over time arises from
spells in different unobserved states.
The techniques will be introduced using examples from locust
behaviour, DNA sequence analysis, and luteinizing hormone levels.
The first two involve discrete outcomes in discrete time and the last
has continuous responses in continuous time.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|