ENS de Lyon, Master IF, HMM for time series classification and filtering¶

Outline, requirement, textbook

Introductory slides (17.8 Mo)

Lesson 1: Mixture Model (MM) and Bayesian Decision (BD) theory¶

• Slides Lesson 1 Chapter 1 & 2 (33.9 Mo)
• Slides Lesson 1 Chapter 3 & 4 (40.3 Mo)
• Lab session statement (date: ???)
  • Pratical work subject (pdf), Skeleton of programs
  • Solution
• Grading: 1h sitting exam, no document, grading weight: 25%, date: ???

Lesson 2: Hidden Markov models (HMM)¶

• Slides Lesson 2 Chapter 1 (44.6 Mo)
• Slides Lesson 2 Chapter 2 (27.1 Mo)
• Python programs (5 Ko)
• Lab session statement (date: ???)
  • Pratical work subject (pdf), Skeleton of programs
  • Solution
• Slides Lesson 2 Chapter 3 - part 1 (43 Mo)
• Slides Lesson 2 Chapter 3 - part 2 (28 Mo)
• Slides from Haoyu LI - Conference : TMC application to quantified-self (3.8 Mo)
• Grading: 2h sitting exam, no document, grading weight: 50%, date: ???

  Please read carefully the following paper, a question will be on it!
  S. Derrode and W. Pieczynski, Unsupervised signal and image segmentation using pairwise Markov chains, IEEE Trans. on Signal Processing, Vol. 52(9), pp. 2477-2489, Septembre 2004.

Lesson 3: Kalman like filtering¶

• Slides Lesson 3 Chapter 1 (78.8 Mo), Pupil tracking video (4.2 Mo)
• Python programs (11 Ko)
• Slides Lesson 3 Chapter 2 (35.9 Mo)
• Lab session statement (date: ???)
  • Pratical work subject (pdf), Skeleton of programs