Arash Tavakoli

Arash Tavakoli


PhD Candidate in Computer Science
(Deep Reinforcement Learning)

Imperial College London

a.tavakoli [at] imperial.ac.uk

twitter github linkedin google scholar

News

Education

  • 2016 – Present:
    • Imperial College London, London, United Kingdom
    • PhD Candidate in Computer Science
    • Focus: Deep Reinforcement Learning
  • 2014 – 2016:
  • 2010 – 2014:
    • University College London, London, United Kingdom
    • MEng in Electrical Engineering (First-Class Honours)
    • Remark: Dean's List Recognition
  • 2012 – 2013:

Publications

Prioritizing Starting States for Reinforcement Learning
Arash Tavakoli*, Vitaly Levdik*, Riashat Islam, Petar Kormushev
In Workshop on Deep Reinforcement Learning, Conference on Neural Information Processing Systems (NeurIPS), 2018.
[arXiv]

Time Limits in Reinforcement Learning
Fabio Pardo, Arash Tavakoli, Vitaly Levdik, Petar Kormushev
In International Conference on Machine Learning (ICML), 2018.
[PMLR Library | arXiv | website | talk]

Action Branching Architectures for Deep Reinforcement Learning
Arash Tavakoli, Fabio Pardo, Petar Kormushev.
In AAAI Conference on Artificial Intelligence (AAAI), 2018.
Press: [Unity ML-Agents blog ("Additional new features")]
[AAAI Library | arXiv | code | Long Talk]

Crowdsourced Coordination through Online Games
Arash Tavakoli, Haig Nalbandian, Nora Ayanian.
In ACM/IEEE International Conference on Human-Robot Interaction (HRI), 2016.
Press: [MIT Technology Review (0:34s)]
[ACM Library | IEEE Xplore Library | preprint]
See also: [ICRA workshop version | game design report]

Seamless Robot Simulation Integration for Education: A Case Study
Wolfgang Hönig, Arash Tavakoli, Nora Ayanian.
In Workshop on Simulation in Robot Programming, IEEE International Conference on Simulation, Modeling, and Programming for Autonomous Robots (SIMPAR), 2016.
[PDF | website | code]