Variational inference mpc for bayesian model-based reinforcement learning
M Okada, T Taniguchi - Conference on robot learning, 2020 - proceedings.mlr.press
… recent studies on model-based reinforcement learning (MBRL)… -art strategy to enhance
learning performance, making MBRLs … PaETS can improve asymptotic performance compared to …
learning performance, making MBRLs … PaETS can improve asymptotic performance compared to …
Adapting user interfaces with model-based reinforcement learning
… approach for adaptive UIs that can improve usability while … success of an adaptation, bandits
use Bayes theorem to update … 20 items are still within reach in the case of menu systems. To …
use Bayes theorem to update … 20 items are still within reach in the case of menu systems. To …
Model-based lifelong reinforcement learning with bayesian exploration
… baseline in the single-task setting within the same amount of interactions. Thus, we let them
… To improve sample efficiency in lifelong RL, our work proposed a model-based lifelong RL …
… To improve sample efficiency in lifelong RL, our work proposed a model-based lifelong RL …
Uncertainty-aware model-based reinforcement learning: Methodology and application in autonomous driving
… Then, a novel uncertainty-aware model-based RL method is … , and improving RL’s learning
efficiency and performance. The … to design a feasible model-based RL under an inaccurate …
efficiency and performance. The … to design a feasible model-based RL under an inaccurate …
On the importance of hyperparameter optimization for model-based reinforcement learning
… need for human expertise and further improve performance. … This allows optimizers to learn
interaction effects within each group. When … Fast bayesian optimization of machine learning …
interaction effects within each group. When … Fast bayesian optimization of machine learning …
Revisiting design choices in offline model-based reinforcement learning
… key hyperparameters using Bayesian Optimization produces … average performance over
our final 10 policy-improvement … datasets under the True Model-Based experiment under the …
our final 10 policy-improvement … datasets under the True Model-Based experiment under the …
Robust model-free reinforcement learning with multi-objective Bayesian optimization
… that are commonly used in model-based control and we discuss how … Below, we formally
introduce them and we explain how to evaluate … We know that increasing values of delay take a …
introduce them and we explain how to evaluate … We know that increasing values of delay take a …
Efficient hyperparameter optimization through model-based reinforcement learning
J Wu, SP Chen, XY Liu - Neurocomputing, 2020 - Elsevier
… optimizing hyperparameters without human interference within a fixed computing budget (…
times) of the model use and improve the efficiency of optimization while ensuring accuracy. …
times) of the model use and improve the efficiency of optimization while ensuring accuracy. …
Efficient model-based reinforcement learning through optimistic policy search and planning
… , we show in Appendix H.3 that under this assumption we can improve the dependence …
Improving sample efficiency is one of the key bottlenecks in applying reinforcement learning to …
Improving sample efficiency is one of the key bottlenecks in applying reinforcement learning to …
Model-based reinforcement learning with value-targeted regression
… main contribution of this paper is a new model-based upper confidence RL algorithm. The
main … . 2017; Agrawal & Jia 2017 in a Bayesian setting), we propose to evaluate models based …
main … . 2017; Agrawal & Jia 2017 in a Bayesian setting), we propose to evaluate models based …