iamrajatroy / Data-Science-Lab

Awesome repository for all my Data Science Experiments

Data-Science-Lab

Awesome repository for all my Data Science Experiments

About

This respository consists of jupyter notebooks which follows a step by step process which is usually followed in a data science experiment.

The goal is to adding new use-cases which solved a particular problem using Machine Learning and I will be using the data which is readily available on Kaggle or any other open source site.

Experiment 1 - Credit Payment Default

Use Machine Learning to predict whether a customer will default in his/her next payment

Check out the code here - Notebook

Experiment 2 - Tensorflow Sequential API and Subclassing

A walkthrough code for comparing different approaches of model training with Tensorflow

Check out the code here - Notebook

Experiment 3 - Computer Vision Model Explanability

A walkthrough code for visualizing CNN localized maps in a given image using Tensorflow

Notebook