Global ETD Search

Interpretable Deep Learning: Beyond Feature-Importance with Concept-based Explanations

Dimanov, BottyUniversity of Cambridge2020

… to improve interpretability. Feature importance explanations are the most popular interpretability approaches. They show the importance of each input feature (e.g., pixel, patch, word vector) to the model’s prediction. However, we hypothesise that feature importance explanations have two main …

cambridge Repository record for Interpretable Deep Learning: Beyond Feature-Importance with Concept-based Explanations (opens in a new tab)