omagdy / malaria-diagnosis

A CNN model that diagnoses whether a microscopic image contains malaria or not

Malaria Diagnosis

Introduction

This project utilizes a CNN to detect whether a colored microscopic image contains malaria or not.

Training

To train the model use the following command:

python main.py -lr 0.01 -bs 200 -ep 100 

where lr is the learning rate, bs is the batch size and ep is the number epochs for training.

Possible Improvements

1. Utilize an optimizer to find the optimum hyperparameters for training the model
2. Create a deeper model to obtain better performance and utilize skip connections to minimize the vanishing gradient problem