Published on Tue Aug 28 2018

Cycle-of-Learning for Autonomous Systems from Human Interaction

Nicholas R. Waytowich, Vinicius G. Goecks, Vernon J. Lawhern

We discuss different types of human-robot interaction paradigms in the context of training end-to-end reinforcement learning algorithms. Two key concepts provided by our Cycle-of-Learning framework are how it handles the integrationof the different human-interaction modalities.

0
0
0
Abstract

We discuss different types of human-robot interaction paradigms in the context of training end-to-end reinforcement learning algorithms. We provide a taxonomy to categorize the types of human interaction and present our Cycle-of-Learning framework for autonomous systems that combines different human-interaction modalities with reinforcement learning. Two key concepts provided by our Cycle-of-Learning framework are how it handles the integration of the different human-interaction modalities (demonstration, intervention, and evaluation) and how to define the switching criteria between them.

Sun Aug 30 2020
Artificial Intelligence
Human-in-the-Loop Methods for Data-Driven and Reinforcement Learning Systems
Recent successes combine reinforcement learning algorithms and deep neural networks. This can be attributed to the fact that current end-to-end reinforcement learning approaches still require thousands or millions of data samples to converge to a satisfactory policy. In real world scenarios and after just a few data samples, humans are able to either provide demonstrated tasks or intervene to prevent catastrophic actions.
0
0
0
Mon Jul 15 2019
Machine Learning
Mutual Reinforcement Learning
Collaborative robots have begun to train humans to achieve complex tasks. A new approach called mutual reinforcement learning (MRL) can lead to successful collaborations. Both humans and autonomous agents act as reinforcement learners in askill transfer scenario.
0
0
0
Wed Jun 09 2021
Machine Learning
PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training
Human-in-the-loop RL methods allow practitioners to instead interactively teach agents through tailored feedback. Such approaches have been challenging to scale since human feedback is very expensive. We present an off-policy, interactive RL algorithm that capitalizes on the strengths of both feedback and
8
38
200
Thu Mar 06 2014
Artificial Intelligence
Active Learning for Autonomous Intelligent Agents: Exploration, Curiosity, and Interaction
Different communities proposed different solutions, that are in many cases, similar and/or complementary. These solutions include active learning, exploration/exploitation, online-learning and social learning. The common aspect of all these approaches is that it is the agent to selects and decides what information to gather.
0
0
0
Wed Oct 09 2019
Machine Learning
Ctrl-Z: Recovering from Instability in Reinforcement Learning
When learning behavior, training data is often generated by the learner. This can result in unstable training dynamics. This problem has particularly important applications in safety-sensitive real-world control tasks. We propose a principled and model-agnostic approach to mitigate the issue of unstable learning dynamics.
0
0
0
Fri Jul 03 2020
Artificial Intelligence
A Conceptual Framework for Externally-influenced Agents: An Assisted Reinforcement Learning Review
The long-term goal of reinforcement learning agents is to be able to perform complex real-world scenarios. There is a general lack of collaboration or interoperability between different approaches using external information. The proposed taxonomy details the relationship between external information source and the learner agent.
0
0
0