Research
Research
Understanding human preferences and predicting human behavior in complex and interactive environments has been a longstanding focus across econometrics, social science, and machine learning, with successful applications in various domains such as resource planning and infrastructure development. Nowadays, the modeling problem becomes more and more complex given the availability of massive amounts of data collected from various channels (mobile phones, online platforms, social networks, etc.), especially when humans now can have access to various sources of information and even interact with AI agents when making decisions. My long-term aim is to address the following questions: (i) How can we efficiently model and predict human behavior in complex, dynamic, and interactive environments? (ii) How can we make the modeling and prediction process adaptive and robust to handle different circumstances of data availability and human behavior constraints? And (iii) how can we make the behavioral models useful for downstream decision-making tasks? These questions pose several new challenges that necessity new methodologies, and my strategies are to address these challenges through methods rooted in Econometrics, Operations Research, and Machine Learning, focusing on applications in transportation behavioral modeling, location planning, revenue management, and game theory.
Theory
Discrete choice theory
Robust/Stochastic optimization
Dynamic programming
Reinforcement/Inverse Reinforcement Learning/Imitation learning
Simulation-based optimization
Applications
Behavior modeling
Transportation planning
Revenue/workforce management
AI/Sercurity Games
Selected Recent Publications
Tran H.*, Mai T., Network-based Representations and Dynamic Discrete Choice Models for Multiple Discrete Choice Analysis. Transportation Research Part B (2024) [Discrete Choice] [Route Choice] [pdf]
Le C.*, Mai T., Constrained Assortment Optimization under the Cross-Nested Logit Model. Production and Operations Management (2024)[Discrete Choice] [Revenue Mangement] [pdf] [link]
Mai T., Bui T.V.*, Nguyen Q.P., Le T.V., Estimation of Recursive Route Choice Models with Incomplete Trip Observations. Transportation Research Part B (2023) [Discrete Choice] [Route Choice] [pdf]
Hoang H*., Mai T., Varakantham P. Imitate the Good and Avoid the Bad: An Incremental Approach to Safe Reinforcement Learning, in Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI2024 - Acceptance rate: 23.75%) . [Imitation Learning][IRL/RL][pdf]
Bui T.V.*, Mai T., Thanh H. Nguyen, Imitating Opponent to Win: Adversarial Policy Imitation Learning in Two-player Competitive Games, in Proceedings of the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS2023 - Full paper - Acceptance rate: 23.3%), [IRL/RL] [Imitation Learning][Competitive Games] [pdf]
Mai T., Sinha A., Choices Are Not Independent: Stackelberg Security Games with Nested Quantal Response Models, in Proceedings of the 36th AAAI Conference on Artificial Intelligence (AAAI2022) (Acceptance rate: 1349/9020 = 15%). [Discrete Choice] [Nested logit] [Security Game] [pdf]
Full list of publications
Journal Publications
Le C.*, Mai T., Constrained Assortment Optimization under the Cross-Nested Logit Model. Production and Operations Management (2024)[Discrete Choice] [Revenue Mangement] [pdf] [link]
Tran H.*, Mai T., Network-based Representations and Dynamic Discrete Choice Models for Multiple Discrete Choice Analysis. Transportation Research Part B (2024) [Discrete Choice] [Route Choice] [pdf]
Ngan H.*, Dam T.T.*, Ta T.A., Mai T. Joint Location and Cost Planning in Maximum Capture Problems under Random Utility Models. Computer and Operations Research (2023). [Discrete Choice] [Facility Location] [pdf]
Dam T.T*., Ta. T.A, Mai T. Robust Maximum Capture Facility Location under Random Utility Models. European Journal of Operational Research (2023). [Discrete Choice] [Facility Location] [pdf]
Mai T., Bui T.V.*, Nguyen Q.P., Le T.V., Estimation of Recursive Route Choice Models with Incomplete Trip Observations. Transportation Research Part B (2023) [Discrete Choice] [Route Choice] [pdf]
Mai T., Frejinger E. Undiscounted Recursive Path Choice Models: Convergence Properties and Algorithms. Transportation Science (2022). [Discrete Choice] [Route Choice] [link] [pdf]
Dam T.T.*, Ta T.A., Mai T., Joint Chance-constrained Simulation-based Staffing Optimization in Multi-skill Call Centers, Journal of Combinatorial Optimization (2022). [Discrete Optimization] [Stochastic Simulation]
Dam T.T.*, Ta T.A., Mai T., Submodularity and Local Search Approaches for Maximum Capture Problems under Generalized Extreme Value Models. European Journal of Operational Research (2021). [Discrete Choice] [Facility Location] [pdf]
Mai T., Yu X., Gao S., Frejinger E., Route choice in a stochastic time-dependent network: the recursive model and solution algorithm. Transportation Research Part B, Volume 151, (2021). [Discrete Choice] [Route Choice] [pdf]
Ta T.A., Mai T., Bastin F., l'Ecuyer P., On a multistage discrete stochastic optimization problem with stochastic constraints and nested sampling, Mathematical Programming (2020). [Stochastic Programming] [Stochastic Simulation] [pdf]
de Moraes Ramos G., Mai T., Daamen W., Frejinger E., Hoogendoorn S. P. Route choice behaviour and travel information in a congested network: Static and dynamic recursive models. Transportation Research Part C: Emerging Technologies (2020) 114, 681-693. [Discrete Choice] [Route Choice] [pdf]
Mai T., Lodi A. A multicut outer-approximation approach for competitive facility location under random utilities. European Journal of Operational Research 284.3 (2020): 874-881. [Discrete Choice] [Facility Location] [pdf]
Bastin F., Liu Y., Cirillo C., Mai T., Transferring Time-Series Discrete Choice to Link-based Route Choice in Space: Estimating Vehicle Type Preference Using Recursive Logit Model , Transportation Research Record (2018). [Discrete Choice]
Mai T., Frejinger E., Fosgerau M., Bastin F., A dynamic programming approach for quickly estimating large network-based MEV models, Transportation Research Part B: Methodological 98 (2017): 179-197. [Discrete Choice] [pdf]
Zimmermann M., Mai T., Frejinger E. Bike route choice modeling using GPS data without choice sets of paths, Transportation Research Part C: Emerging Technologies 75 (2017): 183-196. [Discrete Choice] [Route Choice]
Mai T. , Frejinger, E., Bastin, F. On the similarities between random regret minimization and mother logit: the case of recursive route choice models, Journal of Choice Modeling (2017). [Discrete Choice] [Route Choice]
Mai T., A method of integrating correlation structures for a generalized recursive route choice model, Transportation Research Part B, 93(1):p.146-161, 2016. [Discrete Choice] [Route Choice] [pdf]
Mai T., Fosgereau M., Frejinger E., A nested recursive logit for route choice analysis, Transportation Research Part B, 71(1):p.100-112, (2015). [Discrete Choice] [Route Choice] [pdf]
Mai T., Frejinger, E., Bastin, F., A misspecification test for logit-based route choice models, Economics of Transportation, 4(4):p.215-226, (2015). [Discrete Choice] [Route Choice]
Mai T., Bastin F., Frejinger E., A decomposition method for estimating complex recursive logit-based route choice models, EURO Journal on Transportation and Logistics (2015): 1-23. [Discrete Choice] [Route Choice] [pdf]
Book Chapters
Mai T., Lodi A., An algorithm for assortment optimization under parametric discrete choice models. Fields Institute Communications Series on Data Science and Optimization (2023 - forthcoming). [Discrete Choice] [Revenue Mangement] [pdf]
Conference Proceedings
Mai T., Bose A., Sinha A., Thanh H. Nguyen, Stackelberg Network Interdiction under Boundedly Rational Adversary. Forthcoming in the 33rd International Joint Conference on Artificial Intelligence (IJCAI2024). [Dynamic Discrete Choice] [Behavioral game theory][pdf]
Jiang H.*, Mai T., Varakantham P., Hoang H*, Solving Constrained Reinforcement Learning through Augmented State and Reward Penalties, in Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI2024 - Acceptance rate: 23.75%) [IRL/RL][pdf]
Hoang H*., Mai T., Varakantham P. Imitate the Good and Avoid the Bad: An Incremental Approach to Safe Reinforcement Learning, in Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI2024 - Acceptance rate: 23.75%) . [Imitation Learning][IRL/RL][pdf]
Bui T.V.*, Mai T., Imitation Improvement Learning for Large-scale Capacitated Vehicle Routing Problems, in Proceedings of the 33rd International Conference on Automated Planning and Scheduling (ICAPS 2023) [Combinatorial optimization] [Imitation learning][pdf]
Bui T.V.*, Mai T., Thanh H. Nguyen, Imitating Opponent to Win: Adversarial Policy Imitation Learning in Two-player Competitive Games, in Proceedings of the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS2023 - Full paper - Acceptance rate: 23.3%), [IRL/RL] [Imitation Learning][Competitive Games] [pdf]
Bose A.*, Li T., Sinha A., Mai T., A Fair Incentive Scheme for Community Health Workers, in Proceedings of the 37th AAAI Conference on Artificial Intelligence (AAAI2023 - Acceptance rate: 19.6%) [IRL/RL][Bounded Rationality][AI for Social Impact] [pdf]
Mai T., Sinha A., Safe Delivery of Critical Services in Areas with Volatile Security Situation via a Stackelberg Game Approach, in Proceedings of the 37th AAAI Conference on Artificial Intelligence (AAAI2023 - Acceptance rate: 19.6%) [Discrete Choice] [Security Game] [Workshop presentation at OptLearnMas22 - AAMAS22] [pdf]
Bose A.*, Sinha A., Mai T., Scalable Distributional Robustness in a Class of Non-Convex Optimization with Guarantees, in Proceedings of the 36th Conference on Neural Information Processing Systems (NeurIPS 2022) (Acceptance rate: 25.6%)[Discrete Choice] [Robust Optimization] [pdf]
Ta T.A., Mai T., Bastin F., l'Ecuyer P. A logistic regression and linear programming approach for multi-skill staffing optimization in call centers, in Proceedings of the Winter Simulation Conference 2022, Singapore (WSC2022) Singapore [Discrete Optimization] [Simulation] [pdf]
Mai T., Sinha A., Choices Are Not Independent: Stackelberg Security Games with Nested Quantal Response Models, in Proceedings of the 36th AAAI Conference on Artificial Intelligence (AAAI2022) (Acceptance rate: 1349/9020 = 15%). [Discrete Choice] [Nested logit] [Security Game] [pdf]
Barabonkovy D.*, D'Alonzoy S.*, Pierre J.*, Kondor D., Zhang X., Mai T. (2020). Simulating and evaluating rebalancing strategies for dockless Bike-Sharing Systems, in Proceedings of the Transportation Research Board (TRB) 99th Annual Meeting, January 12--16.
Working/Unpublished Papers
Shao Q., Mai T., and Cheng Shih-Fen. Constrained Pricing in Choice-based Revenue Management. [Discrete Choice] [Revenue Mangement]
Hoang H*., Mai T., Varakantham P. SubIQ: Inverse Soft-Q Learning for Offline Imitation with Suboptimal Demonstrations (submitted) . [Imitation Learning][IRL/RL][pdf]
Pham H.G*, Dam T.T*, Duong N.H.*, Mai T., Ha M.H., Competitive Facility Location under Random Utilities and Routing Constraints (submitted). [Discrete Choice] [Facility Location]
Bui T.V.*, Mai T., Thanh H. Nguyen, Inverse Factorized Q-Learning for Cooperative Multi-agent Imitation Learning (submitted). [IRL/RL] [Imitation Learning][Multiagent Games] [pdf]
Bui T.V.*, Mai T., Thanh H. Nguyen, Unleashing Imitation Learning Potential in Two-Player Competitive Games (submitted). [IRL/RL] [Imitation Learning][Competitive Games] (Extended journal version of our AAMAS23 paper).
Bui T.V.*, Mai T., Thanh H. Nguyen, Mimicking To Dominate: Imitation Learning Strategies for Success in Multiagent Competitive Games (submitted). [IRL/RL] [Imitation Learning][Competitive Games] [pdf]
Duong N.H.*, Mai T., Ta T.A., Binary-Continuous Sum-of-ratios Optimization: Discretization, Approximations, and Convex Reformulations (submitted) [Discrete Choice] [Combinatorial optimization] [pdf]
Bui T.V.*, Mai T., Jaillet P., Weighted Maximum Entropy Inverse Reinforcement Learning. [Behavior Learning] [IRL] [pdf]
Mai T., Jaillet P., Robust pricing under generalized extreme value models. [Discrete Choice] [Revenue Mangement] [pdf]
Mai T., Jaillet P., On the relation between Markov Decision Process frameworks. [IRL/RL] [pdf]
Mai T., Jaillet P., Robust entropy-regularized Markov Decision Processes. [IRL/RL] [pdf]
Mai T., Nguyen Q. P., Low K. H., Jaillet P., Inverse reinforcement learning with missing data. [IRL/RL] [Route Choice] [pdf]
(*) indicate student or research engineer
Others
Mai T. (2016). Dynamic programming approaches for estimating and applying large-scale discrete choice models, Ph.D. thesis, University of Montreal.
Mai T. (2013). Revisiting optimization algorithms for maximum likelihood estimation, Master thesis, University of Montreal.