Bandit: ContextualBinaryBandit

Contextual Bernoulli multi-armed bandit where at least one context feature is active at a time.

Usage

  bandit <- ContextualBinaryBandit$new(weights)

Arguments

weights

numeric matrix; d x k matrix with probabilities of reward for d contextual features per k arms

Methods

new(weights)

generates and initializes a new ContextualBinaryBandit instance.

get_context(t)

argument:

• t: integer, time step t.

returns a named list containing the current d x k dimensional matrix context$X, the number of arms context$k and the number of features context$d.

get_reward(t, context, action)

arguments:

• t: integer, time step t.

• context: list, containing the current context$X (d x k context matrix), context$k (number of arms) and context$d (number of context features) (as set by bandit).

• action: list, containing action$choice (as set by policy).

returns a named list containing reward$reward and, where computable, reward$optimal (used by "oracle" policies and to calculate regret).

See also

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

Bandit subclass examples: ContextualBinaryBandit, ContextualLogitBandit, OfflineReplayEvaluatorBandit

Policy subclass examples: EpsilonGreedyPolicy, ContextualLinTSPolicy

Examples

if (FALSE) {

library(contextual)

horizon            <- 100
sims               <- 100

policy             <- LinUCBDisjointOptimizedPolicy$new(alpha = 0.9)

weights             <- matrix(  c(0.4, 0.2, 0.4,
                                  0.3, 0.4, 0.3,
                                  0.1, 0.8, 0.1),  nrow = 3, ncol = 3, byrow = TRUE)

bandit             <- ContextualBinaryBandit$new(weights = weights)

agent              <- Agent$new(policy,bandit)

history            <- Simulator$new(agent, horizon, sims)$run()

plot(history, type = "cumulative", regret = TRUE)

}