🏦 Smart Loan Risk Predictor (ML + Streamlit)

---

📌 Project Overview

This project is a production-ready Smart Loan Risk Predicton system designed to help financial institutions assess credit risk and predict the probability of loan risk using machine learning.

The solution combines advanced ML models with a high-end Streamlit dashboard, delivering an experience similar to real-world fintech products used by banks and NBFCs.

🔗 GitHub Repo: Link

🌐 Live Demo: Link

---

🎯 Problem Statement

Loan defaults pose significant financial risks for lending institutions.
The goal of this project is to predict the likelihood of a borrower defaulting on a loan using demographic, financial, and loan-specific attributes.

📈 This enables:

• Better credit decisions
• Early risk identification
• Data-driven lending strategies

---

🧠 Machine Learning Approach

✔ Models Used

• XGBoost Classifier
• Bagging Ensemble Technique
• Feature importance–based selection
• Class imbalance handling using sample weighting

✔ Evaluation Metrics

• Accuracy
• Precision
• Recall
• F1-Score
• ROC-AUC (for probability calibration)

---

📊 Dataset Description

The dataset contains borrower-level information with the following features:

Category	Examples
Demographics	Age, Education, MaritalStatus
Financial	Income, CreditScore, DTIRatio
Loan Details	LoanAmount, LoanTerm, InterestRate
Employment	EmploymentType, MonthsEmployed
Target	Default (0 = No, 1 = Yes)

📁 Files:

• train.csv → Training data with target
• test.csv → Test data (no target)
• prediction_submission.csv → Final predictions

---

🖥️ Streamlit Web Application

🔥 Key Features

• Modern fintech-style UI
• Interactive sliders & dropdowns
• Real-time default probability prediction
• Feature importance visualization
• Risk-level interpretation (Low / Medium / High)

🎨 UI Highlights

• Glassmorphism cards
• Gradient theme
• Interactive Plotly charts
• Sidebar navigation

---

🛠️ Tech Stack & Tools

👨‍💻 Programming & ML

📊 Visualization & App

---

📁 Project Structure

Smart-Loan-Risk-Predictor/
│
├── SmartLoanRiskPredictor.ipynb
├── app.py
├── train.csv
├── test.csv
├── prediction_submission.csv
├── requirements.txt
├── README.md
├── .gitignore
└── venv/

---

🚀 How to Run the Project

1️⃣ Clone the Repository

git clone https://github.com/mr-piyushkr/Smart-Loan-Risk-Predictor.git
cd Smart-Loan-Risk-Predictor

---

2️⃣ Create & Activate Virtual Environment

python -m venv venv
venv\Scripts\activate   # Windows

---

3️⃣ Install Dependencies

pip install -r requirements.txt

---

4️⃣ Run Streamlit App

streamlit run app.py

---

🧪 Model Output

• Predicted Probability of default (0–1)
• Risk category: 🟢 Low Risk 🟡 Medium Risk 🔴 High Risk

---

📌 Key Learnings

• End-to-end ML pipeline design
• Handling class imbalance effectively
• Feature engineering & selection
• Ensemble learning with XGBoost
• Deploying ML models using Streamlit
• Designing professional ML dashboards

---

🌐 Future Improvements

• Model monitoring & logging
• API integration (FastAPI)
• Database support
• Cloud deployment (AWS / GCP)

---

📄 License

This project is licensed under the MIT License.

---

👨‍💻 Author

Piyush Kumar
🚀 Data Science & Machine Learning Developer

📬 Let's Connect

🌐 Portfolio  •  💻 GitHub  •  💼 LinkedIn  •  📧 Email