Supervised Machine Learning – Loan Approval Prediction - ML Model training and building

This project focuses on using machine learning techniques to analyze customer financial details and predict whether a loan application is likely to be approved or rejected. A link for the Loan Approval Prediction Dataset from Kaggle.

📌 Project Goal

Build classification models to predict loan approval outcomes using supervised machine learning. The project includes data preprocessing, model training, evaluation, and model selection.

🧠 Techniques Used

• Logistic Regression
• Decision Tree Classifier
• Hyperparameter tuning using GridSearchCV
• Cross-validation
• Evaluation metrics: Accuracy, Precision, Recall, F1-score, ROC AUC
• Data visualization (EDA & Feature Importance)

📊 Dataset

The dataset includes details such as:

• Applicant income
• Loan amount
• Loan term
• CIBIL score
• Asset values
• Education level
• Employment status
• Target variable: loan_status (Approved / Rejected)

✅ Tasks Completed

1. Preprocessing

   • Handling missing values
   • Encoding categorical variables
   • Feature scaling

2. Modeling

   • Trained Logistic Regression and Decision Tree models
   • Performed cross-validation and hyperparameter tuning

3. Evaluation

   • Compared models using key metrics and visualizations
   • Analyzed feature importance

4. Conclusion

   • Decision Tree selected as the final model based on performance

📈 Key Visualizations

• Loan status distribution
• Boxplot: Annual income vs Loan status
• Correlation heatmap
• Feature importance (Decision Tree)

🧾 Final Report

See the full report in the loan_approval_prediction.ipynb notebook or exported PDF.

📂 Structure

Supervised_ML/

• Dataset

  • loan_approval_dataset.csv

• Loan_approval_prediction

  • loan_approval_prediction.ipynb

• Report.pdf

• README.md