Printer-friendly version

Resources 

• Old Mid-term Exams
  
  
  1. Midterm Exam, Spring'07
     
     
  
  
• Viterbi Decoding
  
  
  1. Viterbi Animation
     
     
  2. Viterbi Demo
     
     
  
  
• MCMC, Gibbs Sampling
  
  
  1. Gibbs Sampling with Example
     
     
  2. MCMC,Gibbs Intro
     
     
  
  
• JT Olds' Python Code
  
  
  1. Sampling Algorithms
     
     Rejection Sampling
     Likelihood Weighting
     Gibbs Sampling
     
     which require
     
     bayesnets.py
     bayestools.py
     markovtools.py
     
     To run, for example, the rejection sampling script for 1000 samples, type the following:
     
     ./rejectionsampling.py 1000
     
     
  2. Exact Inference Algorithms
     
     Sum-Product Algorithm
     Max-Product Algorithm
     
     These are standalone, so to run, for example, the sum-product algorithm, type
     
     ./sumproduct.py
     
     These have good printouts to the screen so you can follow the methods. The following two algorithms only print out the final solution, but you can look at the source files for the methods.
     
     Enumeration
     Variable Elimination
     
     
  3. Filtering/Smoothing/Prediction and Viterbi Decoding (Homework 2, Problem 4)
     
     prob4.py
     
     which requires
     
     hmm_tools.py
     hmms.py
     
     Run by typing
     
     ./prob4.py