Interpretable Machine Learning

Title: Interpretable machine learning - what does it actually mean?

Abstract: In recent years, the field of interpretable machine learning has received an increasing amount of attention. In the absence of a well-formed definition of interpretability, a broad range of methods with a correspondingly broad range of outputs (e.g. visualizations, natural language, mathematical equations) have been labeled as interpretation. This has led to considerable confusion about the notion of interpretability. In particular, it is unclear what it means to interpret something, what common threads exist among disparate methods, and how to select an interpretation method for a particular problem/audience.

I'll be discussing a recent submission to PNAS, https://arxiv.org/pdf/1901.04592.pdf, where we take a stab at formally defining interpretable machine learning, introduce our Predictive, Descriptive, Relevant (PDR) framework for evaluating interpretations, and what we see as the major challenges moving forward, all grounded in real-world examples. This is joint work with Chandan Singh, Karl Kumbier, Reza Abbasi-Asl and Bin Yu (my advisor).

Full details about this meeting will be posted here: https://www.benty-fields.com/manage_jc?groupid=191. 

The Berkeley Statistics and Machine Learning Forum meets weekly to discuss current applications across a wide variety of research domains and software methodologies. Register here to view, propose and vote for this group's upcoming discussion topics. All interested members of the UC Berkeley and LBL communities are welcome and encouraged to attend. Questions may be directed to François Lanusse.