Credit Applicant Risk Assessment

Project Summary:

The focus of the project was to use various decision tree models to determining if credit applicants presented a good or bad risk. The dataset was based on German credit data and had 1000 past credit applicants with 30 variables. In this class we used RapidMiner. The analysis and topics covered were thorough including lift, model selection, misclassification costs, and cumulative profit.

Steps Involved:

• Data Cleaning and Exploration
• Descriptive Model to Identify Important Variables for Credit Risk
• Predictive Model to Identify "Good" vs. "Bad" Credit Applicants
• Assessing Performance using Misclassification Costs
• Using a Cumulative Profit Graph to Determine Model Implementation
  
  

File Uploaded:

1. "RashmiMariyappa_CreditRisk_Report"
   The report details the steps and analysis for this project.