Ali Tajer

Research interests

I am broadly interested in the theoretical and algorithmic aspects of machine learning, mathematical statistics, and information theory. Currently, my focus is on causality, representation learning, and bandit algorithms.

Complete list of publications

Curriculum Vitae (last updated December 2024)

Short bio

I received the B.Sc. (2002) and M.Sc. (2004) degrees in Electrical Engineering from Sharif University of Technology, and the M.A. degree in Statistics and the Ph.D. degree (2010) in Electrical Engineering from Columbia University. I was a postdoctoral research associate at Princeton University from 2010 to 2012.

Editorial boards

Associate Editor, IEEE Transactions on Information Theory
Senior Area Editor, IEEE Transactions on Signal Processing

Graduate Students:

I am always on the lookout for Ph.D. students interested in machine learning, information theory, and statistical inference (see this page for more information).

Recent representative publications:

Paper on task-oriented causal representation learning (NeurIPS 2025).
Paper on robot pose estimation via causal representation learning (NeurIPS 2025 Workshop on Embodied World Models).
Paper on tail-sensitive bandits for wireless communications (NeurIPS 2025 workshops on AI4NextG).
Paper on causal representation learning -- linear and general transformations (JMLR 2025).
Paper on preference-centric bandits (arXiv).
Paper on risk-sensitive bandits (AISTATS 2025).
Paper on causal prediction of cascading anomalies in power systems (IEEE Transactions on Power Systems 2025).
Paper on oracle complexity of combinatorial bandits (TMLR 2025).
Manuscript on causal representation learning -- linear transformations (arXiv).
Paper on multi-node causal representation learning (NeurIPS 2024).
Paper on finite-sample causal representation learning (NeurIPS 2024).
Paper on intervention-based learning of DAG mixtures (NeurIPS 2024).
Paper on causal bandits with unknown DAGs and soft interventions (NeurIPS 2024).
Paper on physics-guided RL for power systems (NeurIPS 2024).
Paper on addressing ß-optimality of BAI in stochastic bandits (IEEE Transactions on Information Theory, 2024).
Manuscript on physics-guided RL for blackout mitigation in power systems (IEEE Transactions on Power Systems, 2024).
Paper on best arm identification in restless bandits (IEEE Transactions on Information Theory, 2024).
Paper on the robustness of causal bandits (IEEE Journal on Selected Areas in Information Theory, 2024).
Paper on causal representation learning -- general transformations (AISTATS 2024).
Paper on causal bandits with general models and continuous interventions (AISTATS 2024).
Paper on causal discovery from mixture models (Transactions on Machine Learning Research, 2024).
Paper on causal bandits with linear models (Journal of Machine Learning Research, 2023).
Paper on SPRT-based BAI in stochastic bandits (IEEE Journal on Selected Areas in Information Theory, 2024).
Paper on estimating high-dimensional graphical structures (IEEE Transactions on Information Theory, 2023).
An edited book entitled "Advanced Data Analytics for Power Systems", Cambridge University Press.

Overview paper (arXiv, Entropy) on the broadcast approach in communication networks.

Information Sciences Group

Professor, ECSE and CS

Rensselaer Polytechnic Institute