Interpretable machine learning:

definitions, methods, and applications

Jamie Murdoch, Chandan Singh, Karl Kumbier, Reza Abbasi Asl, Bin Yu

interpretable machine learning: the use of machine-learning models for the extraction of relevant knowledge about domain relationships contained in data

​

Predictive accuracy

Descriptive accuracy

�Relevant

An interpretation is relevant if it provides insight for a particular audience into a chosen domain problem.

​

the degree to which an interpretation method objectively captures the relationships learned by ML models.

Relevancy

Model-based

• sparsity
• simulatability
• modularity
• feature engineering

Post hoc

• dataset-level
• feature importances
• visualization
• trends + outliers
• prediction-level

Future work

• Evaluating desiderata (PDR) of different methods
• measuring descriptive accuracy
• demonstrating relevancy to real-world problems
• new model-based methods
• new post hoc methods
• what should interpretations look like?
• improving predictive accuracy with interpretations