# Why maximum likelihood estimation is different in a logistic regression classifier? [duplicate]

why don't we use the simple linear regression method such as mean square for parameter fitting in logistic regression?