Skip to content

Ridanshi/FinTrust-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 

Repository files navigation

Loan Approval Prediction System

Overview

This project builds a supervised Machine Learning model to predict whether a loan application should be Approved (1) or Rejected (0) based on applicant financial, personal, and credit-related information.

The goal is to automate the loan screening process, reduce manual bias, and improve decision consistency in banking systems.

Objective

  • Predict loan approval status using historical applicant data
  • Reduce financial risk by identifying high-risk applicants
  • Improve efficiency compared to manual verification processes

Dataset Description

Each row represents a loan applicant with features such as:

  • Applicant_Income
  • Coapplicant_Income
  • Employment_Status
  • Age
  • Marital_Status
  • Dependents
  • Credit_Score
  • Existing_Loans
  • DTI_Ratio (Debt-to-Income Ratio)
  • Savings
  • Collateral_Value
  • Loan_Amount
  • Loan_Term
  • Loan_Purpose
  • Property_Area
  • Education_Level
  • Gender
  • Employer_Category

Target Variable:

  • Loan_Approved → 1 (Approved), 0 (Rejected)

Tech Stack

  • Python
  • Pandas
  • NumPy
  • Scikit-learn
  • Matplotlib / Seaborn

Machine Learning Workflow

  1. Data Cleaning & Preprocessing
  2. Handling Missing Values
  3. Encoding Categorical Variables
  4. Feature Scaling (StandardScaler)
  5. Train-Test Split
  6. Model Training (e.g., Logistic Regression / Random Forest)
  7. Model Evaluation using:
    • Accuracy
    • Precision
    • Recall
    • F1-Score
    • Confusion Matrix

Results

The trained classification model successfully predicts loan approval status with strong performance metrics, demonstrating the effectiveness of supervised learning for financial risk assessment.

About

Machine Learning-based Loan Approval Prediction System using supervised classification models.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors