Can we use continuous variables as an input for logistic regression?

regression
logistic

#1

Hi,

I am new to business analytics domain and always refer a table for selecting model methodology like if input and output variables are continuous then I go with Linear regression, if input is categorical and output is continuous then ANOVA and if both are categorical then Logistic regression.

Recently I have read about logistic regression online and it is mentioned there, it can be used for both categorical and continuous predictor variables. Can you please help me to understand it better?

Regards,
Imran


#2

Imran,
You are right with the Linear regression and ANNOVA. However, for a logistic regression you need to have at least one continuous variable to build a reasonable model. Just imagine if you have only categorical variables as a predictor, something like a CART/CHAID will work much better. Obviously your target variable is a categorical variable.
Hope this helps.
Tavish