BootstrapTSPolicy: Policy: Thompson sampling with the online bootstrap

Description Details Usage Arguments References See Also

Description

Bootstrap Thompson Sampling

Details

Bootstrap Thompson Sampling (BTS) is a heuristic method for solving bandit problems which modifies Thompson Sampling (see ThompsonSamplingPolicy) by replacing the posterior distribution used in Thompson sampling by a bootstrap distribution.

Usage

1
policy <- BootstrapTSPolicy(J = 100, a= 1, b = 1)
1

Arguments

new(J = 100, a= 1, b = 1)

Generates a new BootstrapTSPolicy object. Arguments are defined in the Argument section above.

set_parameters()

each policy needs to assign the parameters it wants to keep track of to list self$theta_to_arms that has to be defined in set_parameters()'s body. The parameters defined here can later be accessed by arm index in the following way: theta[[index_of_arm]]$parameter_name

get_action(context)

here, a policy decides which arm to choose, based on the current values of its parameters and, potentially, the current context.

set_reward(reward, context)

in set_reward(reward, context), a policy updates its parameter values based on the reward received, and, potentially, the current context.

References

Eckles, D., & Kaptein, M. (2014). Thompson sampling with the online bootstrap. arXiv preprint arXiv:1410.4009.

Thompson, W. R. (1933). On the likelihood that one unknown probability exceeds another in view of the evidence of two samples. Biometrika, 25(3/4), 285-294.

See Also

Core contextual classes: Bandit, Policy, Simulator, Agent, History, Plot

Bandit subclass examples: BasicBernoulliBandit, ContextualLogitBandit, OfflineReplayEvaluatorBandit

Policy subclass examples: EpsilonGreedyPolicy, ContextualLinTSPolicy


Nth-iteration-labs/contextual documentation built on March 10, 2020, 6:50 a.m.