Style Sampler

Layout Style

Patterns for Boxed Mode

Backgrounds for Boxed Mode

Search News Posts

  • General Inquiries 1-888-555-5555

  • Support 1-888-555-5555

anaplatform Data Consultancy
Credit Risk Analysis

We support your Digital Transformation journey with our analytical finance solutions.

Credit Risk Analysis

Credit Risk Analysis

Credit risk management; It is under the authority and evaluation of the bank that provides the loan to the person. In the loan allocation process, banks decide whether to take a loan after evaluations in accordance with their own decision processes.

The outputs of internal risk rating models, developed using statistical methods on historical data, are included in the relevant lending policies and procedures in order to rate customers for the corporate and commercial loans portfolio using objective criteria.

This plug and play solution enables operations, strategy and transformation teams to understand how anaplatform can be used to understand and explain patterns in a loan application workflow through automated and interactive process mining techniques.

We know that there are models created with different effect and subpopulation analysis. biased data will produce bias estimates. In anaplatform, differential impact analysis measures whether a sensitive group has achieved a positive result at a close rate. that of the advantaged group.

In addition, anaplatform's risk analysis allows users to view results by groups. to weed out unwanted model biases. Both analyzes help to find groups of people who may be treated unfairly or differently by the model to help. teams deliver more responsible and equitable results.

Model Fairness

Teams should dig deeper before suggesting a solution to model bias. How biased a model is can be done by measuring some measure of fairness. anaplatform's model fairness will be designed to help accomplish this task. Depending on the individual use case, different fairness measures may need to be applied. The purpose of the platform is to show users - in a transparent way - various measures of fairness. and the differences between them. From there, users can choose the best one and evaluate the fairness of the current situation.

Four different metrics will be calculated, so let's look at a loan example. Each one needs to be evaluated.

  • Demographic Equality: People across groups have the same chance of getting credit.
  • Equalized Odds: Chances of getting the same loan among people who will not default. The same chance of rejection among people who will default.
  • Equal Opportunity: Among all people who will not default, the chance of getting a loan is the same.
  • Predictive Ratio Parity: Among all creditors, there is the same proportion of people in groups who do not default (success considered equal chance).



Result: