C2W2 Viterbi algorithm Accuracy is 0.9528 and expected is 0.9531

[VibhavGarg]

Hello, for Viterbi algorithm, I am getting accuracy of 0.9528 whereas expected in 0.9531. Any ideas why?

[arvyzukai]

Hi @VibhavGarg

Check this thread and see if that is your problem too (Exercise 7).

[VibhavGarg]

Hi @arvyzukai , The other tests all passed. I had a problem with the for loop index, i fixed that

[VibhavGarg]

Not sure what the issue is, the autograder passed all tests. Thanks

[arvyzukai]

@VibhavGarg , have you passed all the tests but not sure what the problem was?

[VibhavGarg]

@arvyzukai yes

[awebdesigner09]

Same error here. No clue on what the issue is. All other tests passed except accuracy.
I see for loop is already provided. Any help is appreciated:

Tried to see debug with print() statements. But, no luck.

@Deepti_Prasad - tagged from other thread.

[Deepti_Prasad]

yo can DM the code screenshot. click on my name and then message. I am suspecting one of the key function might be used incorrectly

[Deepti_Prasad]

@awebdesigner09

1. Issue in viterbi forward

For code line

#Calculate the probability = 
# best probs of POS tag k, previous word i-1 +  
# log(prob of transition from POS k to POS j) + 
# log(prob that emission of POS j is word i)
                

previous POS tag mentioned in one of the iteration as a list but incorrectly, make sure j is is written as [j]

2. Issue in viterbi backward

For code line

## Step 2 ##
# Find the best POS tags by walking backward through the best_paths
# From the last word in the corpus to the 0th word in the corpus

z is the unique ID which will get you the best pos tag for position i in the corpus,so it should be z[i]

Your compute accuracy code is correct.

Regards
DP

[awebdesigner09]

Thanks for the response. Are you referring to Viterbi forward or backward methods?

[Deepti_Prasad]

My comments mentions both grade cells, so yes. Try to read responses with carefully line by line, it mentions both.

[awebdesigner09]

Thanks for your response. I tried to figure out, unfortunately I did understood/found issue you pointed out for ‘forward’.

And, I have corrected backward by using z[i] as suggested. That matched accuracy and solved the issue.

Thank you so much.