2. Explanations Many methods for explainable ML use the following setup to explain a specific prediction from a complex model: Construct a simpler, easier-to-explain model (e.g., linear regression, decision tree, etc.) that behaves similarly to the complex model for data points near the specific point we're trying to explain. Interpret the simpler model. In this question, we'll try to see if we can come up with an explanation for the worst predictions from each model