This chapter explained the creation of a predictive model. Here, we learned that the process is very iterative. Even though we initially explored data, we did it even more times after its transformation, looking for better corrections that will help to fit data better to a predictive model.
In the first two sections, Assignment and preparation and Data exploration, we described the physical structure of source data and used several techniques, namely T-SQL queries, the SSIS Data Profiling Task, and R scripts, to explore data and to uncover its hidden patterns.
In the Data transformation section, we made several T-SQL transformations and explored data once more to ensure that the transformations did not corrupt the data.
During the last section, we developed training stored procedures with an evaluation of the model quality, and we also created a stored procedure that will...
