Business Problem

Dream Housing Finance company is a growing company that provides home loans in all over US Areas. They first check the eligibility of the customers before customers apply for the application. The company is interested in automating the pre-approval eligibility process by evaluating the customer’s data provided by them on the application. This is important, as this provides the customer their likelihood of getting approved before they apply for the loan. The data is provided by the company to determine the eligibility of the customer so they can target these customers. There are many factors that can determine the eligibility of the customer, such as education level, Income, Credit History, etc.

Techniques

I used many techniques for data preparation. I performed data cleaning as well as data transformation for various predictor variables. After cleaning the data was split into train and test set. Three classification model was used for load prediction Logistic Regression, Decision Tree Model and Random Forest Model. Feature Selection was performed to increase accuracy of the model and parameters were tuned for model evaluation. Lastly, Model selection was done using k-fold cross-validation method.

This project lasted approximately 3-4 weeks. Most of the time was spent to do data cleaning, data transforming and exploratory data analysis of all variables.

Key Skills

Data Cleaning, Data Transformation, Data Visualization, Classification Models, Principal Component Analysis, Feature Engineering