https://github.com/sarveshdhond/credit_risk_model

In this project, a dataset from Kaggle was used to build a machine learning model that predicts credit risk based on multiple applicant attributes such as age, financial history, and loan-related details.
https://github.com/sarveshdhond/credit_risk_model

data-science python scikitlearn-machine-learning

Last synced: 3 months ago
JSON representation

In this project, a dataset from Kaggle was used to build a machine learning model that predicts credit risk based on multiple applicant attributes such as age, financial history, and loan-related details.

• Host: GitHub
• URL: https://github.com/sarveshdhond/credit_risk_model
• Owner: SarveshDhond
• Created: 2025-09-22T16:50:44.000Z (10 months ago)
• Default Branch: main
• Last Pushed: 2025-09-22T19:19:51.000Z (10 months ago)
• Last Synced: 2025-10-30T10:26:06.537Z (9 months ago)
• Topics: data-science, python, scikitlearn-machine-learning
• Language: Jupyter Notebook
• Homepage:
• Size: 499 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# PROJECT BACKGROUND

Credit risk assessment is a crucial part of the financial services industry. Banks and lending institutions need to determine whether a loan applicant is a good risk or a bad risk. Traditionally, this is done through manual evaluation of demographic, financial, and behavioral factors.

In this project, a dataset from Kaggle was used to build a machine learning model that predicts credit risk based on multiple applicant attributes such as age, financial history, and loan-related details.

# DATA STRUCTURE AND INITIAL CHECKS

## DATA STRUCTURE

The dataset includes the following key columns:

Age -> Age of the applicant.

Sex –> Gender of the applicant.

Income Sources –> Number of income streams (e.g., salary, business, investments).

Housing –> Housing status (own, rent, free).

Savings Account –> Balance in savings account.

Chequing Account –> Balance in current account.

Credit Amount –> Loan amount requested.

Duration –> Duration of the loan (in months).

Purpose –> Purpose of the loan (car, education, furniture, etc.).

Risk (Target Variable) –> Whether the applicant is a good risk (1) or bad risk (0).

# INITIAL DATA CHECKS 

- Checked for missing values and handled them appropriately.

- Converted categorical variables (e.g., housing, purpose, sex) into numeric form using encoding techniques.

- Verified class balance in the target variable (good vs. bad risk).

- Checked for feature correlation

- 

# EXECUTIVE SUMMARY

The project workflow involved the following steps:

1. Data Understanding & Cleaning – Performed exploratory data analysis (EDA), identified missing values, and standardized categorical and numerical variables.

2. Feature Engineering – Encoded categorical features and normalized numeric features for better model performance.

3. Model Selection – Tested multiple machine learning models (e.g., Decision Tree, Random Forest, XGBoost) to predict risk.

4. Training & Validation – Split the dataset into training and testing sets to evaluate performance.

5. Hyperparameter Tuning – Used grid search/random search to optimize model parameters.

# MODEL PERFORMANCE

1. Decision Tree Classifier

   - Best parameters -> max_depth=7, min_samples_leaf=1, min_samples_split=2

   - Accuracy -> 0.72

2. Random Forest Classifier

   - Best parameters -> max_depth=7, n_estimators=100

   - Accuracy -> 0.77

  

3. XGboost Classifier

   - Best parameters -> max_depth=7, n_estimators=100

   - Accuracy -> 0.72

# INSIGHTS :

- Features such as credit amount, savings account balance, and duration were strong predictors of credit risk.

- Demographic factors like age and sex also contributed but had less predictive power compared to financial attributes.

- The model shows strong potential for use in credit risk screening systems, assisting banks in making data-driven lending decisions.