CHAMP: Changepoint Detection Using Approximate Model Parameters

Abstract

We introduce CHAMP, an algorithm for online Bayesian changepoint detection in settings where it is di cult or undesirable to integrate over the parameters of candidate models. Rather than requiring integration of the parameters of candidate models as in several other Bayesian approaches, we require only the ability to t model parameters to data segments. This approach greatly simpli es the use of Bayesian changepoint detection, allows it to be used with many more types of models, and improves performance when detecting parameter changes within a single model. Experimental analysis compares CHAMP to another state-of-the-art online Bayesian changepoint detection method.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 2014
Accession Number
ADA605983

Entities

People

  • Andrew G. Barto
  • Christopher G. Atkeson
  • Sarah Osentoski
  • Scott Niekum

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Air Platforms

DTIC Thesaurus Topics

  • Algorithms
  • Bayesian Networks
  • Computational Science
  • Computer Science
  • Computer Vision
  • Detection
  • Distribution Functions
  • Gaussian Distributions
  • Hidden Markov Models
  • Markov Models
  • Models
  • Monte Carlo Method
  • Normal Distribution
  • Probabilistic Models
  • Probability
  • Probability Distributions
  • Standards

Fields of Study

  • Computer science

Readers

  • Astronomy and Astrophysics.
  • Distributed Systems and Data Platform Development
  • Regression Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms