Loan Granting and Status Prediction

This classification Project for the banking domain provides hands-on skills to handle tabular data for banking domain, and build classification models to predict the Loan status of the customer.

Loan granting and status prediction is a classification project for the banking domain. It is used to predict whether a loan will be granted or not. The project is based on data from the UCI Machine Learning Repository.

The project uses a Support Vector Machine (SVM) to learn from the data and predict the status of new loan applications. The SVM is trained on 80% of the data and tested on the remaining 20%. The accuracy of the project is 96.2%.

What will you Learn in the Project classification Project for banking domain?

As part of this project you will learn the following:

1. Doing Exploring Data Analysis for a better understanding of the data
2. Handling tabular data for predictive modelling
3. Visualizing data for a better understanding
4. Understanding corrupted data such as missing values, and outliers and treating it
5. Building linear and ensemble-based models

Tools Used

1. Google Colab for training the model
2. Matplotlib/seaborn library for visualizing the loss & accuracy of the model
3. scikit-learn for model building and evaluation

Prerequisite

1. Working knowledge of scikit-learn library
2. Theoretical understanding of handling missing values and outliers
3. Theoretical understanding of linear regularized models such as (lasso/ridge) and different ensembling learning approaches.
4. Understanding of different classification models evaluation metrics

You will get hint/theory videos for each task to help you better understand the project.

Tasks Performed

As part of this project we will be performing following tasks:

Task-1: Load the data and perform EDA for data understanding

Task-2: Cleaning  and preprocessing of the data

Task-3: Treat outliers and impute null values

Task-4: Prepare the data for model building

Task-5: Build different Machine Learning models and compare the performance