Samuel Horvath

Samuel Horvath

PhD Student

KAUST

Biography

This website will not be updated. Please visit my new website.

I am the last year PhD student in Statistics under the supervision of professor Peter Richtárik at Visual Computing Center KAUST. During my PhD, I interned at Amazon hosted by Cedric Archambeau, at Samsung hosted by Stefanos Laskaridis, Mario Almeida and Nicolas Lane, and Facebook hosted by Michael Rabbat. My research interest is mainly in Non-convex and Convex Optimization for Machine Learning, in particular Federated Learning. Before that, I was an undergrad student in Financial Mathematics at Comenius University. I received the Best Poster Award at Data Science Summer School 2018 at Ecole Polytechnique, and I am a recipient of the Best Paper Award at Federated Learning Workshop NeurIPS 2020.

Interests
  • Federated Learning
  • Stochastic Optimization
  • Deep Learning
Education

  • PhD in Statistics, Summer 2022 (Expected)

    KAUST

  • MS in Statistics, 2018

    KAUST

  • BSc in Financial Mathematics, 2017

    Comenius University

Recent Posts

Paper accepted to AISTATS 2022
International Conference on Artificial Intelligence and Statistics
Sat, Jan 29, 2022 1 min read
Paper accepted to AISTATS 2022
Paper accepted to SIMODS
SIAM Journal on Mathematics of Data Science
Thu, Jan 13, 2022 1 min read
Paper accepted to SIMODS
Paper accepted to DistributedML’21
2nd Workshop on Distributed Machine Learning, co-located with CoNEXT 2021
Tue, Dec 7, 2021 1 min read
Paper accepted to DistributedML’21
One paper accepted to NeurIPS 2021 as Spotlight (Top 3%)
Thirty-fifth Conference on Neural Information Processing Systems
Tue, Sep 28, 2021 1 min read
One paper accepted to NeurIPS 2021 as Spotlight (Top 3%)
Facebook AI Research Montreal
I have joined Facebook AI Research as Research Intern from Aug. 2021 to Dec. 2021.
Sun, Aug 15, 2021 1 min read
Facebook AI Research Montreal
See all posts

Recent Publications

Quickly discover relevant content by filtering publications.
A Field Guide to Federated Optimization
Federated learning and analytics are a distributed approach for collaboratively learning models (or statistics) from decentralized …
PDF Cite
FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout
Federated Learning (FL) has been gaining significant traction across different ML tasks, ranging from vision to keyboard predictions. …
PDF Cite
Hyperparameter Transfer Learning with Adaptive Complexity
Bayesian optimization (BO) is a sample efficient approach to automatically tune the hyperparameters of machine learning models. In …
PDF Cite
Optimal Client Sampling for Federated Learning
It is well understood that client-master communication can be a primary bottleneck in Federated Learning. In this work, we address this …
PDF Cite Code
Lower Bounds and Optimal Algorithms for Personalized Federated Learning
In this work, we consider the optimization formulation of personalized federated learning recently introduced by Hanzely & …
PDF Cite
See all publications

Contact

  • samohorvath11 [at] gmail.com
  • Al-Khawarizmi Applied Mathematics Building (BLDG. #1) | Office Cubicle # 2225-WS18, KAUST, Thuwal, 23955-6900