Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/coder5omkar/lendingclubcasestudy

Performed exploratory data analysis (EDA) on the loan dataset from the Lending Club Case Study to identify the key factors influencing loan defaults.
https://github.com/coder5omkar/lendingclubcasestudy

bivariate-analysis data-science exploratory-data-analysis matplotlib numpy pandas seaborn univariate-analysis

Last synced: 20 days ago
JSON representation

Performed exploratory data analysis (EDA) on the loan dataset from the Lending Club Case Study to identify the key factors influencing loan defaults.

Awesome Lists containing this project

README 

        















## Lending Club Case Study



### Table of Contents

* [General Info](#general-information)

* [Technologies Used](#technologies-used)

* [Conclusions](#conclusions)

* [Acknowledgements](#acknowledgements)



### General Info :



``` 

Lending loans to high-risk applicants is a major cause of financial losses, commonly 

referred to as credit loss. Credit loss occurs when a borrower either refuses to 

repay or defaults on the loan, resulting in the lender losing the funds owed.



The primary goal of this analysis is to identify these high-risk loan applicants, 

enabling the company to minimize the number of such loans and, in turn, reduce the 

total credit loss. This case study focuses on using exploratory data analysis (EDA) 

to pinpoint the factors that contribute to loan default.



The analysis aims to uncover the key variables that significantly predict loan default.

By understanding these factors, the company can improve its risk management and make

more informed decisions regarding its loan portfolio 



```



### Conclusions :



``` 

Key Insights on "Charged-Off" Loans with Higher Default Risk

1.Applicant Type: Renters have a higher likelihood of defaulting.

2.Loan Purpose: Loans taken for debt consolidation show a greater risk of default.

3.Verification Status: Applications marked as "Not Verified" are more likely to default.

4.Loan Term: Terms exceeding 36 months are associated with higher default rates.

5.Funding Range: Funded amounts between $5,000 and $10,000 are at higher risk.

6.Loan Amount: Loans within the $5,000 to $10,000 range have higher chances of default.

7.Installment Range: Monthly installments between $145 and $274 show increased default risk.

8.Debt-to-Income (DTI) Ratio: Ratios between 12% and 18% are linked to higher defaults.

Additional Observations

1.Seasonal Trends: December shows a spike in defaults for disbursed loans.

2.Economic Impact: Financial crises, such as in 2011, correlate with increased charged-off loans.



Based on Bi-Variate Analysis of "Charged-Off" Loans, the Following Applicant Categories Show the Highest Default Risk:

Applicants taking loans for home improvement with an income between $60k - $70k

Applicants with MORTGAGE home ownership and an income between $60k - $70k

Applicants with loan amounts in the range of $30k - $35k who are charged an interest rate of 15% - 17.5%



```



### Technologies Used :

- [Python](https://www.python.org/) version: 3.12.4

- [Numpy](https://numpy.org/) version: 1.26.4

- [Pandas](https://pandas.pydata.org/) version: 2.2.2

- [Seaborn](https://seaborn.pydata.org/) version: 0.13.2

- [Matplotlib](https://matplotlib.org/) version: 3.8.4



### Acknowledgements :



- This project was inspired by UpGrad

- It's part of UpGrad tutorials on Exploratory Data Analysis (EDA) on the learning platform



### Contact :

Created by [@in/omkaramale](https://github.com/coder5omkar)- feel free to contact me!



Developed as part of the Exloratory Data Analysis Module required for Post Graduate Diploma in Machine Learning and AI - IIIT,Bangalore.



This project is open source and available under the [MIT License](https://github.com/coder5om/LendingClubCaseStudy/blob/main/licence.txt).