Brain-Tumor Image Classification

[zestyoreo]

Description

1. Downloaded the Brain Classification dataset from the repo link given and added it to this repo.
   
2. Added code for the Brian_Classification python notebook (named "Brain_Tumour_Classification_200050103.ipynb") and trained the multiclass classifier with a 5 Layer CNN, Adam optimizer for 50 epochs.
   
3. Made use of the ImageDataGenerator module to do data augmentation while training the multi-class classifier.
   
4. Saved the model with best validation accuracy for future training.
   
5. Plotted train and val accuracy and loss function plots for the 50 epochs of training done.

Motivation and Context

It was the task...

Review Time Estimate

• Immediately
• Within one week
• When possible

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have added tests to cover my changes.
• All new and existing tests passed.
• I have read the CONTRIBUTING.md

[Karrthik-Arya]

Mostly the code is fine. There is an overall increase in training accuracy but the validation remains constant. There was also a rise in the training loss in between. The case can be that it might be overfitting or it might not even have converged to the right minima. In either scenario experimenting with the model architecture, making the model deeper or experimenting can help improve the accuracy.