Explaining black boxes with a SMILE

SMILE (Statistical Model-agnostic Interpretability with Local Explanations) is a new method that improves explainability by making use of statistical distance measures.

Machine learning is currently undergoing an explosion in capability, popularity, and sophistication. However, one of the major barriers to widespread acceptance of machine learning (ML) is trustworthiness: most ML models operate as black boxes, their inner workings opaque and mysterious.

Explainability is a key aspect of improving trustworthiness. SMILE builds on previous approaches to provide robust, model-agnostic explanations applicable to a wide range of input data domains.

News

Selected Publications

Explaining black boxes with a SMILE: Statistical Model-agnostic Interpretability with Local Explanations

Koorosh Aslansefat, Mojgan Hashemian, Martin Walker, Mohammed Naveed Akram, Ioannis Sorokos, Yiannis Papadopoulos

IEEE Software, 2023

arXiv PDF Code

We propose SMILE, a new method that builds on previous approaches by making use of statistical distance measures to improve explainability.

Team

Dr Koorosh Aslansefat

Assistant Professor

AI Safety and Explainability

Kuniko Paxton

Fairness

Zeinab Dehghani

GenAI Explainability

Mostafa Anoosha

PGR Student

Privacy-Preserving

Mohadeseh Molapour

GenAI Explainability

Shadie Mohammadi

Geo-Explainability

Louis Donaldson

RL Explainability