Overview
In this project, we aim to analyze loan application data from a finance company specializing in urban loans. Our primary objective is to identify patterns and factors that influence loan defaults. By conducting Exploratory Data Analysis (EDA), we will provide insights to help the company make informed lending decisions, ensuring capable applicants are accepted while minimizing financial risk from potential defaulters.
Business Objectives
-
Mitigate Financial Risk: Understand customer and loan attributes that indicate a high risk of default.
-
Enhance Approval Decisions: Ensure that applicants who can repay are not wrongly denied loans.
-
Improve Lending Strategies: Provide actionable insights to optimize loan approval criteria, potentially adjusting loan amounts or interest rates for risky applicants.
Data Analytics Tasks
- Identify Missing Data and Deal with it Appropriately:
Task: Detect and handle missing data effectively. Tools: Excel functions like COUNT, ISBLANK, IF, AVERAGE, and MEDIAN. Visualization: Bar or column chart to show the proportion of missing values.
- Identify Outliers in the Dataset:
Task: Detect and address outliers in numerical variables. Tools: Excel functions like QUARTILE, IQR, and conditional formatting. Visualization: Box plots or scatter plots to highlight outliers.
- Analyze Data Imbalance:
Task: Determine and assess data imbalance in the dataset. Tools: Excel functions like COUNTIF and SUM. Visualization: Pie chart or bar chart to show the distribution of the target variable.
- Perform Univariate, Segmented Univariate, and Bivariate Analysis:
Task: Conduct analyses to understand variable distributions and relationships. Tools: Excel functions like COUNT, AVERAGE, MEDIAN, filters, sorting, and pivot tables. Visualization: Histograms, bar charts, scatter plots, and heatmaps.
- Identify Top Correlations for Different Scenarios:
Task: Analyze and rank correlations between variables and loan default indicators. Tools: Excel functions like CORREL and correlation matrices. Visualization: Heatmaps to highlight significant correlations.
Conclusion
Through this project, we aim to give the finance company a deeper understanding of the key factors affecting loan defaults. The insights gained will be instrumental in refining their loan approval processes, ultimately contributing to more robust and risk-aware lending strategies.