topic Re: How to predict using Logistic Regression model? in Statistical Procedures

How to predict using Logistic Regression model?

Babloo — Tue, 18 Jul 2017 19:10:47 GMT

Asssume I've built a logistic regression model to predict whether the loan can be given to the customer or not. Assume I've validated and implemented the model.

Now client has given me one record for the customer and asked me to predict whether the loan can given to that customer or not using the model which I built. I'm not sure how to predict the value for one customer using the model and I'm not certain what we need to look in the output.

Re: How to predict using Logistic Regression model?

Reeza — Tue, 18 Jul 2017 19:13:19 GMT

If you validated the model you must have scored the data to test your predictions. It's the same idea, score that record and assign the category.

Re: How to predict using Logistic Regression model?

Babloo — Tue, 18 Jul 2017 19:20:00 GMT

Could you please tell me how can we score the data in logistic regression model?

Re: How to predict using Logistic Regression model?

Rick_SAS — Tue, 18 Jul 2017 19:28:34 GMT

When you run the procedure, you need to store some information so that SAS can later score the data. There are several ways, as explained in the article "Techniques for scoring a regression model in SAS" I recommend the section "The STORE statement and the PLM procedure" because that enables you to score, compute tests/estimates/contrasts, and even produce graphics that visualize the model.