Markov Modulated Poisson Process for Unsupervised Event
Detection in Time Series of Counts
Description
Time-series of count data occur in many different contexts. A
Markov-modulated Poisson process provides a framework for detecting
anomalous events using an unsupervised learning approach.