Handling Prediction Model Errors in Planning for Automated Driving Using POMDPs.
Henrik BeyMoritz SackmannAlexander LangeJörn ThieleckePublished in: ITSC (2021)
Keyphrases
- prediction model
- partially observable markov decision processes
- regression model
- belief state
- belief space
- stochastic domains
- partially observable
- reinforcement learning
- response surface methodology
- predictive state representations
- planning problems
- wavelet neural network
- dynamic programming
- exponential smoothing
- neural network
- bp neural network
- fuzzy neural network
- point based value iteration
- state space
- heuristic search
- software reliability
- dynamical systems
- partially observable markov decision process
- partial observability
- bp network
- domain independent
- ls svm
- optimal policy
- customer churn
- predictive model
- finite state
- markov decision processes
- markov decision problems
- short term