Please note: This PhD seminar will take place in DC 2310 and online.
Kira Selby, PhD candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Pascal Poupart
We propose a general deep architecture for learning functions on multiple permutation-invariant sets. We also show how to generalize this architecture to sets of elements of any dimension by dimension equivariance. We demonstrate that our architecture is a universal approximator of these functions, and show superior results to existing methods on a variety of tasks including counting tasks, alignment tasks, distinguishability tasks and statistical distance measurements. This last task is quite important in Machine Learning. Although our approach is quite general, we demonstrate that it can generate approximate estimates of KL divergence and mutual information that are more accurate than previous techniques that are specifically designed to approximate those statistical distances.
This work has been accepted for publication at UAI-2022: Kira A. Selby, Ahmad Rashid, Ivan Kobyzev, Mehdi Rezagholizadeh, Pascal Poupart. Learning Functions on Multiple Sets using Multi-Set Transformers, International Conference on Uncertainty in Artificial Intelligence (UAI), 2022.