AIAI Seminar - Monday 7 March 2022 - Talk by Hana Chockler

Talk by Hana Chockler

Title:

Causality for Explanations and Other Applications

Abstract:

In this talk, I will (briefly) introduce the theory of actual causality as defined by Halpern and Pearl. This theory turns out to be extremely useful in various areas of computer science due to a good match between the results it produces and our intuition. I will explain the definitions informally using examples from formal verification. I will also introduce the definition of responsibility, which quantifies the definition of causality, again using an example from formal methods (which historically was the first application of responsibility).

We will then look in more detail at the application of causality to explanations of AI decisions and discuss some examples.

Time permitting, we will touch upon the concept of fairness and discrimination in causal models.

The talk is based on a number of papers, and, while not strictly limited to my own research, the topics are quite broad, so I will be mostly talking about my work with different co-authors. The talk is reasonably self-contained.

Bio:

Dr Hana Chockler is a Principal Scientist at a startup company causaLens since 2020. causaLen's mission is to apply causal reasoning to a wide array of domains. Dr Chockler is also a Reader in the Department of Informatics, King’s College London. Prior to joining KCL in 2013, she worked at IBM Research in the formal verification and in software engineering departments. Dr Chockler's research interests span a wide variety of topics, including formal verification and synthesis of hardware and software, and, most recently, causal reasoning applied to AI.