Institute of Communications and Connected Systems


Graph Based Reinforcement Learning

mesh grid mapping over a multidimensional wave patern

1 April 2018

Using graph signal processing to improve the data-efficiency of reinforcement learning algorithms

Funder N/A
Amount N/A

Project website https://www.ee.ucl.ac.uk/

Research theme logos - Sensing, Information and Data Processing
Research topics graph based signal processing | large scale networks | reinforcement learning


Classical reinforcement learning problems suffer from the well-known curse of dimensionality, leading to very slow learning problem in high-dimensional search spaces.

In this project, we aim to overcome this limitation by inferring the structure/geometry of the problem in the learning algorithm in the case of high-dimensional states. 

Possible applications are online control of grid networks or traffic networks.


  • None recorded yet
Conference posters

Graph Signal Processing workshop 2018