I am working on building a Credit Scorecard model. Till now I have performed the following steps:

  1. Data preparation and cleansing
  2. Calculate WOE and IV values
  3. Model Fitting

Now I working on calculating scores for each variable. For this, I need to know 'Target Score', 'Target Odds', and 'pdo'. How these 3 variables are calculated? I have gone through some article but everywhere they assume some Target Score and other variables. One of the article link:

Before this standard approach, I did the following:

  1. Data cleansing, transformation(select transformation with best predictive power), and stepwise logistic regression.

Use the variables selected by the stepwise model and build the final model on it.

I have a couple of question from this:

  1. Which is the best way out of the above 2 approaches?
  2. What if the variables selected based on IV and stepwise are different?
  3. Is it right to compare both the approaches?



1 Answer 1


Target score, target odd, and PDO are not calculated. They are set by the analyst. These 3 values are used to "scale" the raw scores produced by a statistical model. Why is scaling important? Scaling provides some sort of standardization, which is especially useful when you have different scoring models across different portfolios.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.