Difference between revisions of "10-601 HMMs"

From Cohen Courses
Jump to navigationJump to search
Line 11: Line 11:
 
=== Taking home message ===
 
=== Taking home message ===
  
* Why BNs are not enough
+
* Learning with known states: MLE
* The layers of HMMs
+
* Learning with unknown states: EM
* Formal definition of a HMM model
+
* Analyzing protein sequences with HMMs
* Inference with observations
 
* Computing the most likely path
 

Revision as of 10:39, 12 August 2014

This a lecture used in the Syllabus for Machine Learning 10-601 in Fall 2014

Slides

Slides in PDF

Readings

Taking home message

  • Learning with known states: MLE
  • Learning with unknown states: EM
  • Analyzing protein sequences with HMMs