AI Chat AI Image Generator AI Video Text to Speech

Abstraction for Deep Reinforcement Learning

02/10/2022

∙

by Murray Shanahan, et al.

∙

∙

We characterise the problem of abstraction in the context of deep reinforcement learning. Various well established approaches to analogical reasoning and associative memory might be brought to bear on this issue, but they present difficulties because of the need for end-to-end differentiability. We review developments in AI and machine learning that could facilitate their adoption.

Murray Shanahan
27 publications
Melanie Mitchell
19 publications

research

∙ 09/15/2017

Deep Reinforcement Learning for Conversational AI

Deep reinforcement learning is revolutionizing the artificial intelligen...

1 Mahipal Jadeja, et al. ∙

research

∙ 05/19/2022

AIGenC: AI generalisation via creativity

This paper introduces a computational model of creative problem solving ...

0 Corina Catarau-Cotutiu, et al. ∙

research

∙ 06/15/2016

Deep Reinforcement Learning With Macro-Actions

Deep reinforcement learning has been shown to be a powerful framework fo...

0 Ishan P. Durugkar, et al. ∙

research

∙ 07/29/2019

Goal-Driven Sequential Data Abstraction

Automatic data abstraction is an important capability for both benchmark...

3 Umar Riaz Muhammad, et al. ∙

research

∙ 05/11/2021

Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective

Most of the recent deep reinforcement learning advances take an RL-centr...

7 Florin Gogianu, et al. ∙

research

∙ 06/06/2022

Deep Reinforcement Learning for Cybersecurity Threat Detection and Protection: A Review

The cybersecurity threat landscape has lately become overly complex. Thr...

0 Mohit Sewak, et al. ∙

research

∙ 07/20/2018

Learning Heuristics for Automated Reasoning through Deep Reinforcement Learning

We demonstrate how to learn efficient heuristics for automated reasoning...

0 Gil Lederman, et al. ∙