Bayesian optimization for policy search via online-offline experimentation

MaRDI QIDQ5214237zbMATH OpenFDO

Authors Benjamin Letham, Eytan Bakshy

Publication date 7 February 2020

Full work available at URL https://arxiv.org/abs/1904.01049, http://jmlr.csail.mit.edu/papers/v20/18-225.html

zbMATH Keywords

A/B testingBayesian optimizationmulti-fidelity optimizationmulti-task Gaussian processpolicy search

Mathematics Subject Classification ID

Bayesian inference (62F15) Optimal statistical designs (62K05) Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• Constrained Bayesian optimization with noisy experiments
• Using trajectory data to improve Bayesian optimization for reinforcement learning
• Methods and theory for off-line machine learning
• Continuous multi-task Bayesian optimisation with correlation
• scientific article; zbMATH DE number 6433488

Cites work

• A Limited Memory Algorithm for Bound Constrained Optimization
• Computationally efficient convolved multiple output Gaussian processes
• Constrained Bayesian optimization with noisy experiments
• Counterfactual reasoning and learning systems: the example of computational advertising
• Doubly robust policy evaluation and optimization
• Efficient global optimization of expensive black-box functions
• Fast Bayesian hyperparameter optimization on large datasets
• Gaussian processes for machine learning.
• Kernels for vector-valued functions: a review
• Multi-fidelity Gaussian process bandit optimisation
• Multi-fidelity optimization via surrogate modelling
• Scrambling Sobol' and Niederreiter-Xing points
• The origins of kriging
• Using simulation to improve sample-efficiency of Bayesian optimization for bipedal robots
• Using trajectory data to improve Bayesian optimization for reinforcement learning

Cited in

(9)

• Constrained Bayesian optimization with noisy experiments
• Bayesian policy gradient and actor-critic algorithms
• Using trajectory data to improve Bayesian optimization for reinforcement learning
• Tunability: importance of hyperparameters of machine learning algorithms
• Online Bootstrap Inference For Policy Evaluation In Reinforcement Learning
• Offline Multi-Action Policy Learning: Generalization and Optimization
• Neural forecasting of the Italian sovereign bond market with economic news
• Inter-DS: a cost saving algorithm for expensive constrained multi-fidelity blackbox optimization
• Continuous multi-task Bayesian optimisation with correlation

Describes a project that uses

Uses Software

• Matching

This page was built for publication: Bayesian optimization for policy search via online-offline experimentation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5214237)