Predictive Loan Analysis

Overview

Predictive Loan Analysis is a machine learning project aimed at predicting whether a customer is likely to accept a personal loan based on various attributes such as age, income, education, family size, etc. The project involves data exploration, feature selection, and the development of a predictive model using the Random Forest Classifier.

The project provides insights into loan acceptance patterns, helps financial institutions make informed decisions, and automates the prediction process.

Features

• Data Exploration: Visualizing data distributions and relationships.
• Feature Selection: Identifying the most important features for loan acceptance prediction.
• Predictive Modeling: Building and training a Random Forest model.
• Evaluation: Measuring model performance using accuracy and classification metrics.

Technologies Used

• Python: Programming language used for data processing and model development.
• Pandas: For data manipulation and analysis.
• Matplotlib/Seaborn: Data visualization libraries.
• Scikit-learn: Machine learning algorithms and utilities.
• Joblib: Saving and loading the trained model.

Dataset

The dataset contains 5000 records with the following attributes:

• ID: Unique identifier.
• Pin-code: Residential area code.
• Age: Age of the customer.
• Fam members: Number of family members.
• Education: Education level (Under Graduate/Graduate).
• T. Experience: Total professional experience in years.
• Income: Annual income (in local currency).
• Mortgage: Mortgage amount (if any).
• Fixed Deposit: Whether the customer has a fixed deposit (yes/no).
• Demat: Whether the customer has a Demat account (yes/no).
• Net Banking: Whether the customer uses net banking (yes/no).
• Loan: Whether the customer applied for a loan (yes/no).

Project Structure

|-- datasets/
|   |-- loan_dataset.csv           # The dataset for the analysis
|-- loan_analysis.py               # Main Python script for data analysis and model building
|-- loan_prediction_model.pkl      # Saved trained model
|-- requirements.txt               # List of dependencies for the project
|-- README.md                      # Project documentation

How to Install

1. Clone the Repository

git clone https://github.com/your-username/predictive-loan-analysis.git
cd predictive-loan-analysis

2. Set Up a Virtual Environment (Optional but Recommended)

python -m venv venv
source venv/bin/activate  # For Windows: venv\Scripts\activate

3. Install Dependencies

Install all required Python packages using the requirements.txt file:

pip install -r requirements.txt

4. Run the Project

After setting up the environment and installing dependencies, run the main analysis script:

python loan_analysis.py

5. View Results

• The model will be trained and saved as loan_prediction_model.pkl.
• Classification results and model performance will be displayed in the console.

License

This project is licensed under the MIT License. You are free to use, modify, and distribute this project, provided that all copies include the original copyright notice and license. For more details, please refer to the LICENSE file included in this repository.