Speaker:
Professor Oliver Johnson
Institution:
Bristol University, UK
Time:
Wednesday, April 6, 2011 - 4:00pm
Location:
BH 2011
It is well-known that the entropy of an unknown stationary source can be
consistently estimated using an estimator based on the lengths of long
repeated sections of text. I will discuss a method for detecting change
points in data sources based on similar information theory-based
quantities. I will present the results of some simulations and
theoretical results based on properties of the typical set which justify
how successful this method can be