Bank loans analysis (Python)

Project Overview

• We need to know which factors influence to give a loan
• We also need to look for any opportunity to increase the bank’s assets.
• It was found that apart from income, family size is an important factor.
• The libraries used were pandas, numpy, matplotlib, seaborn and scipy.
• You can access the Jupiter Notebook at this LINK

Objectives

Central question:

What is the most important factor to give a personal loan?.

Other questions:

• What is the relationship between education and personal loan?
• What is the distribution of the client population?
• What types of accounts and security do our clients have?
• What is the relationship between income and loans?
• What is the relationship between mortgage and loans?

Data cleaning and preparation

After importing the libraries we prepared the data for analysis.

• Elimination of negative values and useless columns
• Exploration of the correlation of data and outliers: There is a strong correlation between age and experience

• Analysis of the distribution of the data: A large number of outliers were found in income

EDA

The exploratory data analysis was done with the target questions in mind. Several conclusions were obtained which will be presented below.

• Distribution of client education: The most of clients are undergraduates but it appears that education does not have a direct relationship with personal loan.

• Analysis of client account type: The vast majority of clients do not have a security or credit account. It could generate some extra assets for the bank if we find a way to encourage them to have one.

• Personal loan in relation to other factors: It seems that the most important factor for a loan is income, no surprise there. Another really important factor is the number of relatives in the family.

Conclusions

• The most influential values are income and family size.
• The least influential values are age and experience.
• It is advisable to encourage customers to create a credit account and security.
• You can access the code in this LINK