The Explanation Game: A Formal Framework for Interpretable Machine Learning

In Josh Cowls & Jessica Morley (eds.), The 2020 Yearbook of the Digital Ethics Lab. Springer Verlag. pp. 109-143 (2021)
  Copy   BIBTEX

Abstract

We propose a formal framework for interpretable machine learning. Combining elements from statistical learning, causal interventionism, and decision theory, we design an idealised explanation game in which players collaborate to find the best explanation for a given algorithmic prediction. Through an iterative procedure of questions and answers, the players establish a three-dimensional Pareto frontier that describes the optimal trade-offs between explanatory accuracy, simplicity, and relevance. Multiple rounds are played at different levels of abstraction, allowing the players to explore overlapping causal patterns of variable granularity and scope. We characterise the conditions under which such a game is almost surely guaranteed to converge on a optimal explanation surface in polynomial time, and highlight obstacles that will tend to prevent the players from advancing beyond certain explanatory thresholds. The game serves a descriptive and a normative function, establishing a conceptual space in which to analyse and compare existing proposals, as well as design new and improved solutions.

Links

PhilArchive



    Upload a copy of this work     Papers currently archived: 94,549

External links

Setup an account with your affiliations in order to access resources via your University's proxy server

Through your library

Similar books and articles

Mindreading and endogenous beliefs in games.Lauren Larrouy & Guilhem Lecouteux - 2017 - Journal of Economic Methodology 24 (3):318-343.
Optimal stealing time.Andrea Gallice - 2016 - Theory and Decision 80 (3):451-462.

Analytics

Added to PP
2022-03-10

Downloads
22 (#714,231)

6 months
11 (#350,453)

Historical graph of downloads
How can I increase my downloads?

Author Profiles

David Watson
University College London
Luciano Floridi
Yale University

References found in this work

No references found.

Add more references