Classification and Regression with the BigML Dashboard

1.6 Model Summary Report

From a model’s detail view, you can access the model summary report by clicking on the icon highlighted in Figure 1.36 .

The model summary report has two tabs: field importance and summary, explained in the following subsections.

The field importance provides you a measure of how important a field is relative to the other fields. See subsection 1.2.5 for more details.

This summarized view of your model, includes:

Data distribution: for classification problems, percentage of the dataset instances that belong to each of the objective field classes; for regressions, this is the objective field distribution.
Predicted distribution: for classification problems, percentage of predicted instances for each of the objective field classes; for regressions.
Field importance: a measure of how important a field is relative to the other fields (see also subsection 1.2.5 ).
Rules summary: a summary of the rules that the model learned from the dataset This means that for each possible prediction outcome, you can find here a summary of the paths that would produce that prediction.