Question on Logistic Regression

[jefflincoln]

Why do they call it logistic regression when the inverse of the Logit is what is modelled? or am I wrong?

The inverse logit creates the sigmoidal curve where y max is 1 and y min is 0 and x can be from negative to positive infinity.

I feel stupid for asking this. I believe I have it wrong. Any explanation to clear this up for me will be appreciated, Thanks

 

[hlsmith]

Talking with out thinking, but don't you use the Logit to get the log odds and expit to backtranform. So logistic regression generates log odds for coefficients.

 

[noetsi]

Why do they call it regression period at all (the answer has to do with regression to the mean, but that has little to do with modern regression practice or theory).

My point is that I don't think it really matters what they call it, its largely historical accident in many cases.