Linear regression and logistic regression are simple and effective linear techniques for regression and classification tasks, respectively.

Assumptions of linear regression:

  1. A linear relationship exists between the independent and the dependent variables.
  2. The residuals are normally distributed.
  3. The residuals are homoskedastic.
  4. There should be little to no multicollinearity among the independent variables.
  5. The independent variables are not auto-correlated.

Assumptions of logistic regression:

  1. The dependent variable is binary or ordinal, as the case may be.
  2. The observations are independent of each other.
  3. There should be little to no multicollinearity among the independent variables.

Loss funcitons

Linear regression uses a Mean Squared Error (MSE) for the loss function while logistic uses log loss as the loss function.