Safety Assurance with Ensemble-based Uncertainty Estimation and overlapping alternative Predictions in Reinforcement Learning.
Dirk EilersSimon BurtonFelippe Schmoeller da RozaKarsten RoscherPublished in: SafeAI@AAAI (2023)
Keyphrases
- reinforcement learning
- learning algorithm
- function approximation
- machine learning
- multi agent
- partial observability
- neural network
- uncertain data
- markov decision processes
- state space
- estimation algorithm
- model free
- estimation accuracy
- learning tasks
- information security
- estimation error
- random forests
- accurate estimation
- training data