Summary
In this chapter, we learned how to build a linear regression formula and, beyond this, how to visualize the distances between expected values and a model. These distances are input for statistical tests to find out whether the model is good enough to predict new values.
The machine learning workflow to use a model for prediction starts by doing a definition of the target information we expect and data validation, using a chart to see the possible relationships between the variables. We use 80% of the known data to train the model and see whether it returns values that make sense to our experience. With the remaining 20% of the data, we test the model and see whether it fits the data that was not part of the training. Finally, we predict new values. We have to apply our judgment to see whether the regression is working or not.
This knowledge is useful to apply statistical tests that reject the null
hypothesis that the slope of the linear model is equal to zero. A slope...