The dataset contains 27,560 cell images and is divided into two categories of parasitized & uninfected cells. In data preparation, image data labelled from image to array implementation contains RGB values of each image further resized them into a shape of 130 x 130 by taking the mean value of all images.

In model compilation, “binary_crossentropy” and optimizer “adam” was chose, while the main metric was “accuracy”.The calculated test accuracy of the model which comes around 95.32%

Accuracy:

Visualization of Training Loss and test loss:

Classification report:

This Dataset is taken from the official NIH Website: https://ceb.nlm.nih.gov/repositories/malaria-datasets/

Back to Home