Interpretability for Temporal Models

Temporal Prototype Explanation

Method that identifies prototype temporal sequences (most representative) of a prediction class and explains a new prediction by its similarity to these prototypes.

← Tillbaka