Credit Risk Assessment Project

Overview

This project predicts the likelihood of a user defaulting on a loan. It integrates statistical analysis, predictive modeling, and data visualization to provide insights and actionable recommendations for risk mitigation.

---

Pre-requisites

Run the following command:

pip install Flask

---

Running the Program

1. Navigate to the main folder of the project.
2. Execute the following command:

   python project.py runserver --d

---

Tech Stack

• Backend: Python, Flask
• Frontend: HTML, CSS
• Styling Frameworks: Bootstrap, Tailwind
• Data Visualization: Chart.js

---

Program Design and Architecture

Backend

Model.py

• This file contains a class that handles the pre-processing of data for training, trains a decision tree model, processes input data from users, tests the model, and provides the probability of a user defaulting on a loan.
• Algorithm:
  • Encodes categorical columns using label encoding.
  • Splits data into training and testing sets.
  • Trains a decision tree classifier.
  • Processes user input and predicts default probabilities.
• Input: Dataset (credit.csv) and user input.
• Output: Default probability.

DataStats.py

• Provides statistical insights and preprocessing for the dataset.
• Algorithm:
  • Handles missing data in numerical and categorical columns.
  • Computes descriptive statistics (mean, median, mode, etc.).
  • Aggregates demographic statistics by default status.
  • Generates risk mitigation suggestions.
• Input: Dataset or user profile dictionary.
• Output: Statistical insights and risk suggestions.

Chart.py

• Generates bar charts for categorical data distributions, with user-specific values highlighted.
• Algorithm:
  • Reads the dataset and identifies categorical columns.
  • Calculates value frequencies and generates chart configurations.
• Input: Dataset and optional user input.
• Output: List of chart configurations.

project.py

• This is the main script that initializes the Flask application and renders the necessary templates.
• Algorithm:
  • Presents a form for user input.
  • Processes user input and computes default risk using Model.py.
  • Generates charts and insights using Chart.py and DataStats.py.
  • Displays results, including predictions and visualizations.
• Input: User-provided data and dataset.
• Output: Web-based results page.

---

Frontend

index.html

• This template displays a form for users to input data and allows them to submit the form.
• Features:
  • Dropdowns and numeric fields for various loan-related attributes.
  • Responsive design using Bootstrap and Tailwind.
  • Tooltips for user guidance.
• Output: Form data submitted to the Flask backend.

results.html

• This template shows the results, indicating whether the user is likely to default on a loan. Dashboards and some statistics will also be displayed
• Features:
  • Risk probability, insights, and demographic comparisons.
  • Interactive charts for data visualization.
  • Bootstrap-based design for a professional look.
• Output: Risk assessment results and recommendations.

---

Dataset

• File: credit.csv
• Description: Contains historical financial and demographic data used to train the decision tree model. This CSV file contains the dataset used for training the model.

---