In this chapter, we learned how to build a linear regression model, check for violations in the model assumptions, fix the multicollinearity problem, and finally how to find the best model. Here, we were aided by two important assumptions: the output being a continuous variable, and the normality assumption for the errors. The linear regression model provides the best footing for the general regression problems. However, when the output variable is discrete, binary, or multi-category data, the linear regression model lets us down. This is not actually a let down, as it was never intended to solve this class of problem. Thus, our next chapter will focus on the problem of regression models for binary data.