Vardis Kandiros
I am a Graduate Student in the Theory Group at MIT EECS, fortunate to be advised by Costis Daskalakis.
I obtained a Diploma in Electrical and Computer Engineering from National Techincal University of Athens in 2018, where I wrote my diploma thesis under the wise supervision of Dimitris Fotakis.
My research interests are in Machine Learning Theory, Theoretical Computer Science and Causal Inference, as well as their intersections with High Dimensional Probability and Statistics. In particular, I seek to develop Algorithms and Methods that operate using network/dependent data.
Publications
Preprints and Working Papers-
Estimating Ising Models in Total Variation Distance
Constantinos Daskalakis, Vardis Kandiros, Rui Yao[BibTeX]
- The Conflict Graph Design: Estimating Causal Effects under Arbitrary Neighborhood Interference [BibTeX]
-
Learning Gaussian DAG Models without Condition Number Bounds
Constantinos Daskalakis, Vardis Kandiros, Rui YaoICML 2025 [BibTeX]
-
On Sampling from Ising Models with Spectral Constraints
Andreas Galanis, Alkis Kalavasis, Anthimos Vardis KandirosAPPROX/RANDOM 2024 [BibTeX]
-
Learning Hard-Constrained Models with One Sample
Andreas Galanis, Alkis Kalavasis, Anthimos Vardis KandirosSODA 2024 [BibTeX]
- Learning and Testing Latent-Tree Ising Models Efficiently COLT 2023 [BibTeX]
-
EM's Convergence in Gaussian Latent Tree Models
Yuval Dagan, Constantinos Daskalakis, Anthimos Vardis KandirosCOLT 2022 [BibTeX]
- Statistical Estimation from Dependent Data ICML 2021 [BibTeX]
- Learning Ising models from one or multiple samples STOC 2021 [BibTeX]
-
Node-Max-Cut and the Complexity of Equilibrium in Linear Weighted Congestion Games
Dimitris Fotakis, Vardis Kandiros, Thanasis Lianeas, Nikos Mouzakis, Panagiotis Patsilinakos, Stratis SkoulakisICALP 2020 [BibTeX]
- Opinion Dynamics with Limited Information WINE 2018 [BibTeX]