Skip to content

Instantly share code, notes, and snippets.

@sid-sahani

sid-sahani/Performance Scores

Created September 24, 2016 10:35
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save sid-sahani/bf777c7cc86d729127798967256e77ed to your computer and use it in GitHub Desktop.
Save sid-sahani/bf777c7cc86d729127798967256e77ed to your computer and use it in GitHub Desktop.
Download ZIP
Hello folks, I had a doubt concerning my CNN model's Results. In the 4th epoch I get these scores:
Raw
Performance Scores
Epoch 4:
Examples labeled as 0 classified by model as 0: 14 times
Examples labeled as 0 classified by model as 1: 19 times
Examples labeled as 1 classified by model as 0: 4 times
Examples labeled as 1 classified by model as 1: 29 times
==========================Scores=====================================
Accuracy: 0.6515
Precision: 0.691
Recall: 0.6515
F1 Score: 0.6707
====================================================================
Epoch 5 :
Examples labeled as 0 classified by model as 0: 11 times
Examples labeled as 0 classified by model as 1: 22 times
Examples labeled as 1 classified by model as 0: 3 times
Examples labeled as 1 classified by model as 1: 30 times
==========================Scores=====================================
Accuracy: 0.6212
Precision: 0.6813
Recall: 0.6212
F1 Score: 0.6499
=====================================================================
@sid-sahani
Copy link
Author

While progressing through "Epochs", I always saw Results getting an increase, this is the first case I'm seeing where I get a decline also. Does that depict anything unusual in the Data, or it illustrates some fault in my Models' hyperparamters ? Cutting it short, this isn't something supernatural ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment