In previous chapters, we learned to use different regression models to analyze different types of data. We have therefore fully understood the concept that proposes a regression algorithm for each event, which means that data is not all equal and for each data collection there is a regression algorithm that allows extracting knowledge. Numeric data with many predictors must be treated differently from the data that has only one predictor. Just as, different tools have to be adopted in the presence of categorical data, as well as when we handle data with dichotomous responses. We can safely assert that there is a more suited regression algorithm for each type of data and that a predictive analysis of the data in our possession is crucial to addressing our search for the most suitable algorithm.
Based on what we have said, in this chapter we...
