rajatkhatri7 / humanproteinpredictor

HumanProteinPredictor

Overview

This is a classification model which classify the different proteins present in a image of human blood cell.

Description

As part of the ZeroTOGan course, this was the kaggle compeetation. As general to train the model there given images of human protein samples. We convert these images to grey scale and the give the batches of images to the deep Neural Network which classify the different proteins in the images. There are 9 classes of Protein present in the images.

Technology Stack

The key libraries i used to train this model are :

0.For Neural Network Pytorch

1.For cross validation ScikitLearn

2.For Datframe and other things Pandas

3.For working with arrays(pixel values)Numpy

4.For dataVisualisationMatlotlib