Bank Loan Patterns With Data Science
Introduction of Data Science in the Financial Industry
Data Science is wildly adapted in financial industry. Nowadays consumers can take loans through different financing platforms banks, financial firms etc. Bank enables to browse new consumer loan applications containing the applicant’s age, gender, income, employment status, number of dependents and other self-reported personal information, in order to make determinations as to which condition will leads to approved loans.
The goal was to use basic information about customers for exploration data analysis to the process of performing initial investigations on data to discover patterns, spot outliers, and test hypotheses with summary statistics and graphical representations.
Dataset
The dataset is of a Bank that deals with granting home loans in urban, semi-urban, and rural areas. It consists of the following variables: ID, Person Age, Gender, Regional Area, Person’s Income, Self Employed Status, Married Status, Number of Dependents, Education Status, Loan Approved Status.
Factors affecting the loan acceptance decision
Person’s Income: A higher income increases the likelihood of receiving a loan.
Age: Young age person will have better chance of loan approval.
Regional Area: Living place increase trust level to provide loan.
Employment Status: Employed person has good chance to loan approval.
Education: Higher qualifications increases one's chances of landing a high-paying job and in result will earn more money.
Here is a list of other factors: https://money.usnews.com/loans/articles/beyond-credit-scores-factors-that-affect-a-loan-application
Loan Analysis Process
Loading the Data sets
- Importing Libraries - libraries like pandas, numpy, seaboarn, matplotlib
- Loading and show the dataset - loading bank loan data set file
Exploring the Train Data sets
- Check the dataset by rows and columns - know the structure of data set
- See the features in the Dataset - find the Data Type and number of observations in each column
Analyzing Variables in the Data set
- 69% of the loan applications in the data got approved while 31% applications were rejected.
- Loan Status bar chart
Analyzing the categorical variables in the data
- In the data set, 50 percent of the applicants are male, and 50 percent are females.
- Most of the applicants are from Inner City area (45%), followed by Town area (28%), Rural areas (16%), Suburban areas (10%).
- 65% of the applicants are Self Employed
- 46% of the applicants are married.
- 75% of the applicants are graduates.
- 69% of the applicants have loan approved status.
- The Average of applicant’s income is $27936.07
- The Average applicant's age of loan approval is between 40 year to 50 year
- 43% of the applicants have zero dependents.
Conclusion
The loan is analyzed in this project on basis of applicant's different aspects and it is also extends to approval rate applications. I found that the living area of applicant is important with individual's employment status for loan approval. In addition, education of individual is a good indicator of the application pool of the approvals. Another beneficial factor is marriage status, married applicants have a higher chance of loan approval. By this marriage status benefit, another supportive object is number of dependents on the applicant. Less dependents is a good and supportive factor in loan approval success.
The skills I demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
About Author
