What question did this study set out to answer?

The study aims to improve how robots learn from human explanations by using filters for advice and warnings and policies derived from non-specific sentences.

November 14, 2016IEEE Transactions on Cognitive and Developmental Systems

Learning From Explanations Using Sentiment and Advice in RL

View Full Paper

Ask AI

Bookmark

Authors

SKSamantha KreningThe Ohio State University BHBrent HarrisonUniversity of Kentucky KFKaren M. FeighGeorgia Institute of Technology

Discussion

Member takes

Implication

Randomized trial examines how sentiment and advice improve reinforcement learning for robots, suggesting effective human-robot interaction methods.

Key Points

The study aims to improve how robots learn from human explanations by using filters for advice and warnings and policies derived from non-specific sentences.
Developed a software agent to play Mario Bros. using human-like explanations.
Employed sentiment analysis to categorize explanations into actionable advice and warnings.
Implemented object-focused advice to guide agent actions based on environmental context.
Sentiment filtering achieved approximately 85% accuracy in identifying useful advice.
Object-focused advice enhanced the agent's learning compared to using no advice.
The method of human interaction significantly impacted the quality of the advice provided.

Cite This Study

Krening et al. (2016) studied this question.

synapsesocial.com/papers/6a0fab202badbc352afe845c https://doi.org/10.1109/tcds.2016.2628365

View Full Paper

Ask AI

Bookmark

Implication

Randomized trial examines how sentiment and advice improve reinforcement learning for robots, suggesting effective human-robot interaction methods.

Learning From Explanations Using Sentiment and Advice in RL

Discussion

Member takes

Implication

Key Points

Cite This Study

Learning From Explanations Using Sentiment and Advice in RL

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider