I have three research areas: (a) semantics and logics for probabilistic systems and languages (b) theoretical aspects of machine learning (c) quantum information theory. I work on approximation of continuous-state systems and associated metrics and logics. I am working on a quantitative extension of equational logic which allows one to carry out approximate reasoning equationally. I have been working on Stone duality for Markov processes and completeness theorems for Markovian logics. I am also interested in duality for automata and using it for minimization. Recently I have begun working on diffusion and similar continuous-time Markov processes . I have not been active in quantum information theory for a while but I have just started working on it again.

I was selected as the 2021 Milner Lecturer at the University of Edinburgh. The slides are here.

The editorial boards on which I serve are accessible through Links above.

Recent Talks

From bisimulation to representation ...
The Milner Lecture, University of Edinburgh, 30th September 2021.
Representation learning via metrics. Bologna, May 2021.
Distributional analysis of .... MPI, June 2021.
A categorical view of conditional expectation UCR April 2020
Approximation by averaging PI June 2020
More Talks

Recent Conferences and Workshops

CSL 2020 PC member
AAAI 2020 reviewer
MFPS 2019 PC Member
LICS 2019 PC Member
ICML 2019 reviewer
UAI 2019 reviewer

> More Conferences



E-mail: prakash@cs.mcgill.ca
Office: McConnell ENGG North Wing 105
Office Phone: +1 514 3987074,FAX: 3983883
My postal address is here; do not use my office number.

My Fall schedule is here.
Please check my Google cal link below for up-to-date information on my weekly schedule.

Photos My Google calendar  

Publications conferences and workshopsTalks Students Teaching RL Lab Links curricullum