Uncertainty-Aware Action Advising for Deep Reinforcement Learning Agents | AISC

preview_player
Показать описание

Speaker: Matthew Taylor, Felipe Leno da Silva; Discussion Facilitator: Susan Shu Chang

Motivation:
Although Reinforcement Learning (RL) has been one of the most successful approaches for learning in sequential decision making problems, the sample-complexity of RL techniques still represents a major challenge for practical applications. To combat this challenge, whenever a competent policy (e.g., either a legacy system or a human demonstrator) is available, the agent could leverage samples from this policy (advice) to improve sample-efficiency.

In this work, we propose Requesting Confidence-Moderated Policy advice (RCMP), an action-advising framework where the agent asks for advice when its epistemic uncertainty is high for a certain state. RCMP takes into account that the advice is limited and might be suboptimal. Our empirical evaluations show that RCMP performs better than Importance Advising, not receiving advice, and receiving it at random states in Gridworld and Atari Pong scenarios.
  1. LLMs Explained - Aggregate Intellect - AI.SCIENCE
Рекомендации по теме
Uncertainty Aware Action 0:17:03

Uncertainty Aware Action Advising for Deep Reinforcement Learning Agents - Paper Explained!

Uncertainty-Aware Action Advising 0:59:25

Uncertainty-Aware Action Advising for Deep Reinforcement Learning Agents | AISC

Uncertainty-Aware Score Distribution 0:04:56

Uncertainty-Aware Score Distribution Learning for Action Quality Assessment

Training Uncertainty-Aware Classifiers 0:06:39

Training Uncertainty-Aware Classifiers with Conformalized Deep Learning

Uncertainty-Aware Robust Adaptive 0:15:51

Uncertainty-Aware Robust Adaptive Video Streaming with Bayesian Neural Network and Model Predicti...

Uncertainty-Aware Deep Ensembles 0:13:28

Uncertainty-Aware Deep Ensembles for Explainable Time Series Prediction (Kristoffer Wickstrøm, UiT)

A Review of 0:47:18

A Review of Uncertainty for Deep Reinforcement Learning: High Level Overview

CVPR 2021: Learning 0:04:59

CVPR 2021: Learning Probabilistic Ordinal Embeddings for Uncertainty-Aware Regression

A Review of 0:05:24

A Review of Uncertainty for Deep Reinforcement Learning

16. Uncertainty-Aware Time-to-Event 0:02:54

16. Uncertainty-Aware Time-to-Event Prediction using Deep Kernel Accelerated Failure Time Models

Uncertainty-Aware Reinforcement Learning 0:04:14

Uncertainty-Aware Reinforcement Learning For UAV Collision Avoidance

ADSAI 2022 | 0:36:52

ADSAI 2022 | André Martins: Towards Explainable and Uncertainty Aware NLP

Uncertainty Weighted Offline 0:04:48

Uncertainty Weighted Offline Reinforcement Learning (UWAC)

Deep Reinforcement Learning 0:11:49

Deep Reinforcement Learning | Decision Making Under Uncertainty using POMDPs.jl

[TRO 2024] Safe 0:04:11

[TRO 2024] Safe Multiagent Motion Planning Under Uncertainty for Drones Using Filtered Reinforcem...

Rigorous Uncertainty Quantification 0:32:21

Rigorous Uncertainty Quantification for Off-policy Evaluation in Reinforcement Learning: a Variation

Model-Predictive Policy Learning 0:21:22

Model-Predictive Policy Learning with Uncertainty Regularization for Driving in Dense Traffic

Certain Uncertainty: Next 0:55:53

Certain Uncertainty: Next Generation Water Planning for Deep Uncertainty

Shaofeng Zou University 0:44:25

Shaofeng Zou University of Buffalo Robust Reinforcement Learning Under Model Uncertainty

Uncertainty-driven Imagination for 0:02:16

Uncertainty-driven Imagination for Continuous Deep Reinforcement Learning (CoRL 2017)

HARL@ICDL2021: Invited Talk 0:30:57

HARL@ICDL2021: Invited Talk 12 - Action Advising for Accelerating MARL (Filipe Leno Da Silva)

Deep Robust Reinforcement 0:30:45

Deep Robust Reinforcement Learning and Regularization

Motivational short video 0:00:56

Motivational short video - How to succeed - cartoon

Zero-Shot Reinforcement Learning 0:02:57

Zero-Shot Reinforcement Learning on Graphs for Autonomous Exploration Under Uncertainty