This is a good example of modeling imbalance data in classification problem.

Problem Statement -

This is an example of simplified classification model. Objective of this exercise is to predict if we can approve car loans for some applicants. Train data provided with TARGET variable 1 or 0. We have to run and check performance on TEST data

Contents - Exploratory Data Analysis

• Looking at Data
• Plot and visualze Categorial Variables

Corelation and plot Neumerical variables

• Missing Value Stats Phase 1 Modeling

• Simple way of missing value Imputation

• Model preperation

• Random Forest - grid search, hyperparameter tuning and evaluation of best model by cross validation

• XGboost - hyperparameter tuning,random search and evaluation of best model by cross validation

• Conclusion from Phase 1 Phase 2 Modeling

• Missing value imputation by KNN

• Random Forest

• XGboost

• Is there any improvement ?

• Implementation on Test

• Conclusion and Future Work