1. Your Python code: def BaumWelch(N,A,B,Pi,T) :
1. Your Python code: def BaumWelch(N,A,B,Pi,T) :
return [A,B,Pi] # Trained Model
Where:
N: Number of model states A: Transition probability matrix
B: Output probability matrix Pi: Initial state probability matrix
T: Training data Return re-estimated A, B and Pi matrices
Notes:
1. Your code must be original work – That is: it must not be a copy of code found online.
2. Submit your python code in a separate file named “viterbi_
3. As Outlook filters out executable files, submit your work in a .zip file
2. A 2 page write-up describing the Baum-Welch algorithm and your implementation of it
Note: Be sure to list all references
3. Your hand calculations for the first observation vector
4. A sample run for each of the three observation vectors
Note: Your sample run for the 1st two observation vectors must match your hand calculations.
Please Note:
1. Your code must be original work – That is: it must not be a copy of code found online.
2. All references must be cited.
3. No scans of hand-written calculations for any part of the assignment will be accepted.
4.The sample runs may be screen-grabbed.
Leave a Reply