Loan Approval

Problem statement

Estimating whether the loan application will be approved or not. This is a binary classification problem.

Approach

I wanted to do a project to optimize decision tree ID3 estimator. Checking the maximum performance that can be achieved from a single tree.

Steps

Cleaning and missing data treatment
Performing exploratory data anaysis
Feture selection
Hyperparameter tuning using GridsearchCV
Evaluating metrics

Improvements

Using ensemble techniques like Random Forest and adaboost
Using PCA and LDA
Cluster analysis for exploratory data analysis