MODEL DESIGN

 

Before we explore specific supervised learning algorithms, it’s useful to pause and take a high-level look at the full procedure of building a machine learning model. This involves reviewing a number of steps examined in the preceding chapters as well as new methods including evaluate and predict. These 10 steps take place inside your development environment and follow a relatively fixed sequence. Once you are familiar with this framework, you will find it easy to design your own machine learning models from start to finish.

 

Figure 22: An overview of designing a machine learning model

 

1) Import libraries

Given the Python interpreter works from top to bottom through your code, it’s vital to import libraries before calling any of their specific functions. If you attempt to create a heatmap or pairplot without first importing Seaborn and Matplolib, the Python interpreter won’t be able to process your request.

The libraries...