About Me
I’m a Research Assistant in the RLAI lab , supervised by Professor Richard Sutton , working on both theoretical and practical aspects of the Alberta Plan for AI. My current research focuses on designing efficient learning algorithms in reinforcement learning and deep learning, with a strong emphasis on rigorous theoretical guarantees. In addition to practical algorithm design, my expertise lies in machine learning theory and theoretical computer science, particularly in the computational and statistical aspects of learning algorithms. During my MSc at the University of Alberta, I was fortunate to be supervised by Dr. Xiaoqi Tan and supported as a Teaching Assistant and Research Assistant through SODA Lab and Amii funding. As part of my thesis, I had the privilege of collaborating with Dr. Gellért Weise from Google DeepMind, working on computational hardness results in reinforcement learning.
Publications and Preprints
- Computational Hardness of Reinforcement Learning under Partial qπ Realizability
Shayan Karimi and Xiaoqi Tan
(Accepted in NeurIPS 2025 - Poster) - ETGL-DDPG: A Deep Deterministic Policy Gradient Algorithm for Sparse Reward Continuous Control
Ehsan Futuhi, Shayan Karimi, Chao Gao and Martin Müller
(Accepted to TMLR Journal) - On Distributed Algorithms for Minimum Dominating Set Problem
Sharareh Alipoure, Ehsan Futuhi, Shayan Karimi
(pre-print in Arxiv) - Solving Nurikabe with Monte-Carlo Tree Search
Ehsan Futuhi and Shayan Karimi
(Published in ICCIA)
Technical Reports
- Analyzing Twitter text messages related to Covid-19
Shayan Karimi and Ehsan Futuhi - Review of Query Complexity Bounds for Planning in MDPs under Linear Realizability of the Optimal Value Function
Shayan Karimi and Samuel Robertson - Empirical Study of Differential Q-learning on Continuing Control Tasks
(Course Project) - Level Blending with Evolutionary Algorithm in Latent Space of VQ-VAE
(Course Project)