Risk Model Validation

This project is focused on validating risk models by analyzing delivery rates and their statistical properties. It includes data processing, variance calculations, t-tests, and visualizations to assess the performance of different models and technologies.

Features

• Overall Results Display: Summarizes total offered volume, expected value, and delivery rates.
• Delivery Rate Analysis: Calculates the difference between issued and model delivery rates.
• Variance Calculation: Computes variances of delivery rates and checks for significant differences.
• T-Test for Statistical Significance: Performs a t-test to compare delivery rates and assess whether differences are statistically significant.
• Visualization: Plots delivery rate differences and trends by technology category.

Statistical Analysis

• Variance Calculation:

  • Variances of CORE DEL RATE and ISSUED DEL RATE are computed.
  • A comparison is made to check if the variances are significantly different.

• T-Test:

  • A t-test is performed to determine if the mean delivery rates (CORE DEL RATE vs. ISSUED DEL RATE) are statistically different.
  • This helps validate whether the model delivery rate aligns with the issued delivery rate.

How to Use

1. Ensure the required Python libraries are installed:

   • pandas
   • matplotlib
   • scipy (for t-tests)

2. Ensure that the validation.csv dataset is in the Data subfolder.

3. Run the notebook to:

   • View overall results.
   • Analyze delivery rates, variances, and t-test results.
   • Visualize the data.