How can I generate probabilities from a logistic regression?

I have 5 continuous variables and I want to generate probabilities for each one in terms of a binary outcome variable. For instance, given a set of cholesterol levels and past 10-year morbidity (binary), how would I generate the probability of dying given a cholesterol of 300? I know logistic and probit regressions can give you a measure of the significance of predictor variables. How I can I generate the probabilities.

Thanks.

Answer 1

[Jonathan LeSage]

If you have the Statistics Toolbox, you can fit a logistic regression with mnrfit function and then estimate probabilities using the mnrval function. Here is a link to an example in the documentation to get you started:

• http://www.mathworks.com/help/stats/mnrval.html#btpd2uk-2

Hope this helps!

[Ed]

Thanks. That helps. I had seen that example before but was thrown by a typo in the example code.