This project aims to analyse loan approval pattern using Exploratory Data Analysis (EDA) on a loan application dataset.This analysis provide data driven insights of applicant profile which will help bank and other organisation to take proper future decision regarding loan approval.
The dataset include the following features: loan_ID: This include applicant loan id
Gender: whether applicant is male or female
married: whether applicant is married or not
Dependents:number of member depends on applicant
Education:whether applicant is graduate or not
Self_employed: whether applicant is employed or not
Applicantincome: it include applicant income
CoApplicantincome:it include Coapplicant income
Loan amount : amount of loan applicant applied for
Loan amount term:
Credit history: whether applicant has good(1) credit history or not(0)
Property area: whether applicant live in rural urban or semiurban
loan status: whether loan is approved (1) or not(0)
1.Data preparation
The dataset was cleaned and pre -processed , including handling missing value, outlier detection .
2.Exploratory Data Analysis
Using various library like tidyr,dplyr and ggplot2 to plot histogram,boxplot, bargraph ,scatterplot and also using pipe operator and group by function to sort out data relevantly.
Gender: male applicants are more in number than female candidates though both have almost similar approval rates hence gender is not a strong factor.
Martial status: married applicant are more in number than non married applicant also former has higher approval status
Dependents:applicants with no dependents is more to apply for loan than applicants with higher dependents, also applicant with fewer dependents have higher approval rate.
Education: Graduates have higher approval rate compared to non-graduates
self employed : self employed are less in number to apply for loan than those who are not.
The income of majority of the applicant is ranging from (2,500-5,000) per months.
Most approvals cluster around moderate income of (2000-10000) and loan amount 100-200 high loan amount are less frequently approved regardless of income.
1.priotize applicants with a solid credit history they are less risky .for customers with no history history offer small starter loans or credit builder products to bring them into the credit system.
2.Use a debt to income (DTI) ratio policy .Define clear cutoffs (e.g.loan amount should not exceed 40%of total monthly income).This reduces the risk of defaults.
3.partner with universities ,professionaltraining centers,and companies to provide loan schemes to educated and employed individuals.
4.introduce sticter eligibillity checks(or collateral requirements) for applicants with more dependents.
5.semiurbans regions represent low-risk ,high potentials markets banks can focus more on marketing and product and development here .in rural areas aproval criteria may need to be adjusted or banks should offer secured loans (e.g gold ,land or asset backed loans)
6.offer tiered loan products (e.g. ,small tickets with collateral checks, and large loans with strict eligibility)