Hands-On Ensemble Learning with Python

Understanding interpretability

By employing a large number of models, interpretability is greatly reduced. For example, a single decision tree can easily explain how it produced a prediction, by simply following the decisions made at each node. On the other hand, it is difficult to interpret why an ensemble of 1,000 trees predicted a single value. Moreover, depending on the ensemble method, there may be more to explain than the prediction process itself. How and why did the ensemble choose to train these specific models. Why did it not choose to train other models? Why did it not choose to train more models?

When the model's results are to be presented to an audience, especially a not-so-highly-technical audience, simpler but more easily ...