Reviewing and understanding model details
In the last chapter, we created several models for different projects. DataRobot creates 10 to 20 models in a project, and it would be very onerous to look at and analyze the details of all of these models. You do not have to review each of these models, and it is common to review only the top few models before making a final selection. We will now look at the leaderboard for models in the Automobile Example 2 project and select the top model, as illustrated in the following screenshot:
Figure 7.1 – Model information
In the preceding screenshot, we selected the Model Info tab within the Describe tab to get a view of how large the model is and the expected time it takes to create predictions. This information is useful in real-time applications that are time-sensitive and need to score thousands of transactions quickly. Let's now go to the Feature Impact tab within the Understand tab, as shown in the following...
