Brain-Tumor-Classification (98% Accurate)

In this project, I used a residual neural-network on brain scan data to create a model that, given a brain scan image, can predict whether or not the image has a brain tumor. The image below shows the a sample of the images in the dataset used.

Concepts used in this project:

Deep Learning

• Convolutional Neural Networks
• Residual Neural Networks

Data Visualization:

• Matplotlib

How to recreate the results in this project:

• Download the dataset used in this project from here
• put the dataset in a folder called Brain Tumor Dataset and put it in the same directory as the files in this repository
• Download all the dependencies namely: Tensorflow, matplotlib, pandas and numpy
• Open the Resnet.ipynb file and run the cells.

Potential Improvements to this project

• Setting a learning rate decay
• Adding a regularization such as a drop out layer after a convolution block or an identity block

HAPPY CODING!!!