My Interests
- Reinforcement learning (RL) theory
- Robustness to adversarial attacks in RL
- Preference-based RL
Hello!
My name is Andi Nika, and I am a PhD student in the Machine Teaching and Multi-agent Systems groups at MPI-SWS, Saarbrücken, Germany.
I completed my undergraduate degree in Mathematics at the University of Tirana in 2015. Following that, I took a year-long hiatus to explore my artistic passions, participating in The Voice of Albania and reaching the final nights. Afterward, I returned to academia, pursuing graduate degrees in Mathematics and Electrical and Electronics Engineering at Bilkent University, where I developed a strong interest in research in Machine Learning.
More Information (C.V.)Publications and Preprints
-
► Policy Teaching via Data Poisoning in Learning from Human Preferences
A. Nika, J. Nöther, D. Mandal, P. Kamalaruban, A. Singla, G. Radanović
To appear in AISTATS, 2025. -
► Corruption-Robust Offline Reinforcement
Learning from Human Feedback
D. Mandal, A. Nika, P. Kamalaruban, A. Singla, G. Radanović
To appear in AISTATS, 2025. (Oral Presentation) -
► Defending Against Unknown Corrupted
Agents: Reinforcement Learning of Adversarially Robust Nash Equilibria
A. Nika, J. Nöther, A. Singla, G. Radanović
In ICML, 2024 Workshop on Foundations of Reinforcement Learning and Control. -
► Reward Model Learning vs. Direct Policy
Optimization: A Comparative Analysis of Learning from Human Preferences
A. Nika, D. Mandal, P. Kamalaruban, G. Tzannetos, A. Singla, G. Radanović
In ICML, 2024. -
► Corruption-Robust Offline Two-Player
Zero-Sum Markov Games
A. Nika, D. Mandal, A. Singla, G. Radanović
In AISTATS, 2024. -
► Online Defense
Strategies for Reinforcement Learning Against Adaptive Reward Poisoning
A. Nika, A. Singla, G. Radanović
In AISTATS, 2023. -
► Contextual Combinatorial
Volatile Multi-armed Bandits with Adaptive Discretization
A. Nika, S. Elahi, C. Tekin
In AISTATS, 2020. -
► Multi-user Small Base Station
Association via Contextual Combinatorial Volatile Bandits
M. A. Qureshi*, A. Nika*, C. Tekin
(Equal Contribution)
In IEEE Transactions on Communications, 2021. -
► Online Context-aware
Task Assignment in Mobile Crowdsourcing via Adaptive Discretization
S. Elahi*, A. Nika*, C. Tekin
(Equal Contribution)
In IEEE Transactions on Network Science and Engineering, 2023. -
► Pareto Active Learning with Gaussian
Processes and Adaptive Discretization
A. Nika, K. Bozgan, S. Elahi, C. Ararat, C. Tekin
arXiv preprint, 2021. -
► Contextual Combinatorial Bandits with
Changing Action Sets via Gaussian Processes
A. Nika, S. Elahi, C. Tekin
arXiv preprint, 2021.
* Equal contribution