Loan Status Prediction Model

Project Overview

This project focuses on predicting loan approval status using machine learning techniques. The dataset contains various features related to loan applicants, and the goal is to classify whether a loan will be approved or not.

Dataset

• The dataset used is Loan_status.csv.
• It consists of 614 rows and 13 columns.
• Key features include applicant's income, loan amount, credit history, and property area.

Data Preprocessing

1. Checked for null values and removed rows with missing data.
2. Converted categorical values into numerical values:
   • Loan_Status: 'Y' to 1, 'N' to 0
   • Dependents: '3+' to 4
3. Split the data into features (X) and target (Y).
4. Defined categorical and numerical columns for preprocessing.

Model Training

1. Used StandardScaler for numerical feature scaling.
2. Applied OneHotEncoder for categorical feature encoding.
3. Built an SVM model with a linear kernel.
4. Created a pipeline for data preprocessing and model training.

Model Evaluation

• Training Accuracy: 81.02%
• Testing Accuracy: 79.17%

Model Deployment

• The trained model is saved as loan_status_model.pkl using the pickle module.

Libraries Used

• pandas
• numpy
• matplotlib
• seaborn
• scikit-learn
• pickle

Usage

1. Load the trained model from loan_status_model.pkl.
2. Provide applicant data for prediction.
3. Get loan approval status (1 for approved, 0 for not approved).

Conclusion

This loan status prediction model can help financial institutions automate loan approval decisions based on applicant details, improving efficiency and accuracy.