Ryan Kortvelesy

Ryan Kortvelesy

Machine Learning Research Scientist

University of Cambridge

Biography

I am Ryan Kortvelesy, a PhD Student in the Prorok Lab at the University of Cambridge. My research focuses on the intersection between machine learning and mathematics. Pursuing analytical solutions to fundamental problems in machine learning, I hope to develop new theoretical methods that can be applied to real-world problems. Currently, my work delves into the domains of multi-agent reinforcement learning and graph neural networks (or transformers). One application of particular interest to me is that of cooperative multi-agent systems—the task of training a set of agents (e.g. robots) to work together to achieve some emergent macroscopic behaviour.

Interests
  • Reinforcement Learning
  • Self-Supervsied Learning
  • Transformers / Graph Neural Networks
  • Cooperative Multi-Agent Systems
  • Mathematics
Education
  • PhD in Computer Science, 2024

    University of Cambridge

  • BSE in Electrical Engineering, 2019

    University of Pennsylvania

Experience

 
 
 
 
 
Amazon
Software Engineer
Amazon
June 2019 – August 2019 Seattle, WA
Designed an internal tool for bug detection and reporting.
 
 
 
 
 
JHU Applied Physics Lab, NASA
Intern
JHU Applied Physics Lab, NASA
June 2018 – August 2018 Laurel, MD
Developed an algorithm for scene reconstruction from Lidar, automated landing zone selection, and minimum fuel-cost trajectories for the Dragonfly project.
 
 
 
 
 
Mathworks
Software Engineering Intern
Mathworks
June 2017 – August 2017 Natick MA
Created an algorithm for real-time data compression, optimised data structures for MATLAB speed increases, and formulated an approach for Simulink anomaly detection using a graph fourier transform.

Education

 
 
 
 
 
PhD in Computer Science
University of Cambridge
PhD in Computer Science
October 2019 – February 2024 Cambridge, UK
  • Graph Neural Networks for Multi-Agent Learning
  • Supervisor: Amanda Prorok
  • Prorok Lab
 
 
 
 
 
BSE in Electrical Engineering
University of Pennsylvania
BSE in Electrical Engineering
September 2016 – May 2019 Philadelphia, PA
  • Minors in Computer Science and Mathematics
  • Graduated summa cum laude

Recent Publications

All Publications
(2023). Generalised $f$-Mean Aggregation for Graph Neural Networks. NeurIPS.

PDF Cite Code Slides

(2023). Reinforcement Learning with Fast and Forgetful Memory. NeurIPS.

PDF Cite Code Slides

(2023). POPGym: Benchmarking Partially Observable Reinforcement Learning. ICLR.

PDF Cite Code

(2022). Fixed Integral Neural Networks. arXiv.

PDF Cite Code

Contact