Fetching the paper…

Reading the bibliography…

2019

Modularization of End-to-End Learning: Case Study in Arcade Games

Melnik, Andrew, Fleer, Sascha, Schilling, Malte et al.

Original
PDF
Post on X

Understand

Complex environments and tasks pose a difficult problem for holistic end-to-end learning approaches.

Decomposition of an environment into interacting controllable and non-controllable objects allows supervised learning for non-controllable objects and universal value function approximator learning for controllable objects.
Such decomposition should lead to a shorter learning time and better generalisation capability.
Here, we consider arcade-game environments as sets of interacting objects (controllable, non-controllable) and propose a set of functional modules that are specialized on mastering different types of interactions in a broad range of environments.

Modularization of End-to-End Learning: Case Study in Arcade Games · Around

Built on

“Neural reuse: A fundamental organizational principle of the brain”
Michael Anderson · 2010
Earlier work this paper cites.
“Towards more human-like concept learning in machines: Compositionality, causality, and learning-to-learn”, 2014
Brenden Lake · 2014
Earlier work this paper cites.
“Human-level control through deep reinforcement learning”
Volodymyr Mnih et al · 2015
Earlier work this paper cites.
“Universal value function approximators”
Tom Schaul, Daniel Horgan, Karol Gregor and David Silver · 2015
Earlier work this paper cites.
“Openai gym”
Original
Greg Brockman et al · 2016
Earlier work this paper cites.
“A compositional object-based approach to learning physical dynamics”
Original
Michael Chang, Tomer Ullman, Antonio Torralba and Joshua Tenenbaum · 2016
Earlier work this paper cites.

Similar

“Hierarchical deep reinforcement learning: Integrating temporal abstraction and intrinsic motivation”
Tejas Kulkarni, Karthik Narasimhan, Ardavan Saeedi and Josh Tenenbaum · 2016
Cited alongside, same era.
“Hindsight experience replay”
Marcin Andrychowicz et al · 2017
Cited alongside, same era.
“Metacontrol for adaptive imagination-based optimization”
Original
Jessica Hamrick et al · 2017
Cited alongside, same era.
“Neuroscience-inspired artificial intelligence”
Demis Hassabis, Dharshan Kumaran, Christopher Summerfield and Matthew Botvinick · 2017
Cited alongside, same era.
“Rainbow: Combining improvements in deep reinforcement learning”
Original
Matteo Hessel et al · 2017
Cited alongside, same era.

Then

“Schema networks: Zero-shot transfer with a generative causal model of intuitive physics”
Original
Ken Kansky et al · 2017
Later among the works it cites.
“Learning to Run challenge solutions: Adapting reinforcement learning methods for neuromusculoskeletal environments”
Original
Łukasz Kidziński et al · 2018
Later among the works it cites.
“An Empirical Model of Large-Batch Training”
Original
Sam McCandlish, Jared Kaplan, Dario Amodei and OpenAI Team · 2018
Later among the works it cites.
“The World as an External Memory: The Price of Saccades in a Sensorimotor Task”
Andrew Melnik, Felix Schüler, Constantin Rothkopf and Peter König · 2018
Later among the works it cites.
“An Approach to Hierarchical Deep Reinforcement Learning for a Decentralized Walking Control Architecture”
Original

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Malte Schilling and Andrew Melnik · 2018

Later among the works it cites.