Why Linear Regression is not suitable for Classification

                                 Linear Regression vs Logistic Regression 

Why Linear Regression is not suitable for Classification ?

Problem 1: Linear regression model is sensitive to outlier or imbalance data.

                     

Note: Threshold of Feature should not change in predicting algorithm

                    

Problem 2: Extending Linear regression line will give values greater than 1 and below 0.

                    In classification problem values greater than 1 or less than 0 does not represent  anything. Hence model interpretation becomes extremely challenging.

                  
Problem 3: In linear regression model predicted value is continuous , not probabilistic.

Problem 4: Linear regression assumes that error terms are normally distributed, in case of binary classification, this assumption does not hold true.

Problem 5: Linear regression assumes that variance of random errors is constant, which does not hold true for Logistic Regression.