C3W3_Assignment excersise 04 Siamese model evaluation

[Eleonora_Kulberkyte]

Hi, now I am struggling with the excersise 04. Again, I think I do everything right, but I get the following results for accuracy and confusion matrix:

Accuracy 0.81064453125
Confusion matrix:
[[ 931 5451]
[1008 2850]]

Instead of

Accuracy ~0.725
Confusion matrix:
[[4876 1506]
[1300 2558]]

Is there a way, I could debug each element of this excersise?

Best,
Eleonora Kulberkyte

[Deepti_Prasad]

C3W3 Assignment function classify -> unmatch expected output

Hi @skinx.learning Looking at the error message it is visible that you use pred variable instead of y_pred Cheers P.S. there are multiple previous posts related to the classify function (link1, link2, link3 and others) that could offer you hints to solve your problem.

Try to use search tool with similar exercise name, you will find many threads address your issue, helping you debug.

if this doesn’t help, then send screenshot of your grade cell codes in question by personal DM.

Do not post codes here.

Regards
DP

[Deepti_Prasad]

hi @Eleonora_Kulberkyte

your cosine similarity requires correction, the below link should help you understand the error

C3W3 NLP - Ex 2 - Struggling with data types

The hint is a bit misleading, even I was stuck because of the same, arvyzukai NLP mentor guided me through So based on the instruction previous shared, axis should be 1 For the mask exclude positives, your code is correct. Next thing, check if the score codes is recalled as (v2, v1, transpose_b=true), note v2 is placed before v1

Also please make sure your axis values are correct as per instructions mentioned, I am not saying it’s incorrect but go through the instructions before the grade cell once.

Regards
DP

[Eleonora_Kulberkyte]

Alright, I got it right. The problem was the difference in matrix multiplication and pair-wise multiplication. I should have noticed those before! Thanks for your help!