Logistic Regression

Wed, 05 Dec 2018 07:26:51 +0530

In a classification setting, logistic regression models the probability of a response $Y$ belonging to a particaular category. A simple linear regression can not be used for classification as the output of a linear regression can have a range that goes from $-\infty$ to $\infty$ (we need to find the values in the range [0, 1]). Instead, we can transform the output of linear regression such that the output is confined in the range [0, 1]. For this, a logistic function which is given below can be used. \[p(X) = \frac{e^{\beta_0 + \beta X}}{1 + e^{\beta_0 + \beta X}}\]

Manipulating the above function and taking the log, we get

Logistic Regression on Amit Rajan

Logistic Regression