Daniel Vial
About
I'm currently with the Office of Data Science at the SEC, where I use AI and other advanced analytical techniques to help protect investors and maintain fair markets.
I was previously a postdoc working with Sanjay Shakkottai at UT Austin and R. Srikant at UIUC, and a Ph.D. student working with Vijay Subramanian at Michigan. This page archives my research from those two positions.
You can find me on LinkedIn and Google Scholar, or get in touch via the following permanent emails: dvial@umich.edu, daniel.p.vial@gmail.com.
Papers
2023
R. Chawla, D. Vial, S. Shakkottai, R. Srikant, Collaborative Multi-Agent Heterogeneous Multi-Armed Bandits, ICML
D. Vial, S. Shakkottai, R. Srikant, Robust multi-agent bandits over undirected graphs, ACM SIGMETRICS
2022
D. Vial, S. Sanghavi, S. Shakkottai, R. Srikant, Minimax regret for cascading bandits, NeurIPS (selected as oral)
D. Vial, A. Parulekar, S. Shakkottai, R. Srikant, Regret bounds for stochastic shortest path problems with linear function approximation, ICML
D. Vial, A. Parulekar, S. Shakkottai, R. Srikant, Improved algorithms for misspecified linear Markov decision processes, AISTATS
D. Vial, V. Subramanian, Local non-Bayesian social learning with stubborn agents, IEEE TCNS (extended version of EC '19 paper)
2021
D. Vial, S. Shakkottai, R. Srikant, Robust multi-agent multi-armed bandits, ACM MobiHoc (best paper award runner-up)
D. Vial, V. Subramanian, Empirical policy evaluation with supergraphs, IEEE JSAIT (special issue on reinforcement learning)
2020
D. Vial, Structural results and applications for perturbed Markov chains, Ph.D. dissertation
2019
D. Vial, V. Subramanian, A structural result for Personalized PageRank and its algorithmic consequences, ACM SIGMETRICS
D. Vial, V. Subramanian, Local non-Bayesian social learning with stubborn agents, ACM EC
D. Vial, V. Subramanian, On the role of clustering in Personalized PageRank estimation, ACM TOMPECS
Preprints
D. Vial, S. Shakkottai, R. Srikant, One-bit feedback is sufficient for upper confidence bound policies
D. Vial, V. Subramanian, Restart perturbations for reversible Markov chains: trichotomy and pre-cutoff equivalence
Honors
2022 NeurIPS Conference - paper selected as oral-equivalent (the top ≈2% of submissions)
2021 Mobihoc Conference -- best paper runner-up (the top ≈2% of submissions)
2018 Midwest Machine Learning Symposium -- poster prize (the top ≈6% of submissions)