what_if_mortgage_analysis

Mortgage Dataset analysis using Tensorflow and What-If Tool on Google Cloud Platform

Challenge posed by Google:

In this notebook, you will use mortgage data from NY in 2017 to create two binary classifiers to determine if a mortgage applicant will be granted a loan.

You will train a classifier on two datasets. One is trained on the complete dataset, while the other is trained on a subset of the dataset where 90% of the female applicants that were granted a loan were removed from the training data (so the training data has 90% less females that were granted loans).

You will then compare the two models using the What-If Tool.

Both models are trained on their respective datasets and then are to be compared using the What-If Tool.

Tensorflow v2.2 Documentation:

https://www.tensorflow.org/api_docs/python/tf

What-If Tool Documentation:

https://pair-code.github.io/what-if-tool/walkthrough.html

https://github.com/pair-code/what-if-tool

Deploying models to AI Platform

https://cloud.google.com/ai-platform/prediction/docs/deploying-models#gcloud

Video Guide for the Challenge

https://youtu.be/h4NYU7WHdds