Change Point Detection Using Expectation Maximization Approach

• Author / Creator

  Keshavarz, Marziyeh

• Data analysis plays an important role in system modeling, monitoring and optimization.
  Among those data analysis techniques, change point detection has been widely applied in
  various areas including chemical process, climate monitoring, examination of gene expres-
  sions and quality control in the manufacturing industry, etc. In this thesis, an Expectation
  Maximization (EM) algorithm is proposed to detect the time instants at which data prop-
  erties are subject to change. This method performs eciently especially in missing data
  problem or when directly maximizing the likelihood is dicult. The change point detection
  problem is solved under various scenarios including univariate and multivariate data, known
  and unknown covariance. The problem is also extended to changing covariance in the case
  of multivariate data analysis. Moreover, using Bayesian inference method these problems
  are solved and the results are compared with EM. The results show that in terms of com-
  putation, due to some iterations involved in EM algorithm, it has higher computation but
  the convergence is fast.

• Subjects / Keywords

  • Expectation Maximization
  • Change point
  • Bayesian

• Graduation date

  Fall 2013

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3197X

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.