Man pages for contextual
Simulation and Analysis of Contextual Multi-Armed Bandit Policies

AgentAgent
BanditBandit: Superclass
BasicBernoulliBanditBandit: BasicBernoulliBandit
BasicGaussianBanditBandit: BasicGaussianBandit
BootstrapTSPolicyPolicy: Thompson sampling with the online bootstrap
cliprClip vectors
ContextualBernoulliBanditBandit: Naive Contextual Bernouilli Bandit
ContextualBinaryBanditBandit: ContextualBinaryBandit
ContextualEpochGreedyPolicyPolicy: A Time and Space Efficient Algorithm for Contextual...
ContextualEpsilonGreedyPolicyPolicy: ContextualEpsilonGreedyPolicy with unique linear...
ContextualHybridBanditBandit: ContextualHybridBandit
ContextualLinearBanditBandit: ContextualLinearBandit
ContextualLinTSPolicyPolicy: Linear Thompson Sampling with unique linear models
ContextualLogitBanditBandit: ContextualLogitBandit
ContextualLogitBTSPolicyPolicy: ContextualLogitBTSPolicy
ContextualPrecachingBanditBandit: ContextualPrecachingBandit
ContextualTSProbitPolicyPolicy: ContextualTSProbitPolicy
ContextualWheelBanditBandit: ContextualWheelBandit
ContinuumBanditBandit: ContinuumBandit
data_table_factors_to_numericConvert all factor columns in data.table to numeric
dec-setDecrement
EpsilonFirstPolicyPolicy: Epsilon First
EpsilonGreedyPolicyPolicy: Epsilon Greedy
Exp3PolicyPolicy: Exp3
FixedPolicyPolicy: Fixed Arm
formatted_difftimeFormat difftime objects
get_arm_contextReturn context vector of an arm
get_full_contextGet full context matrix over all arms
GittinsBrezziLaiPolicyPolicy: Gittins Approximation algorithm for choosing arms in...
GradientPolicyPolicy: Gradient
HistoryHistory
inc-setIncrement
indOn-the-fly indicator function for use in formulae
invInverse from Choleski (or QR) Decomposition.
invgammaThe Inverse Gamma Distribution
invlogitInverse Logit Function
is_rstudioCheck if in RStudio
LifPolicyPolicy: Continuum Bandit Policy with Lock-in Feedback
LinUCBDisjointOptimizedPolicyPolicy: LinUCB with unique linear models
LinUCBDisjointPolicyPolicy: LinUCB with unique linear models
LinUCBGeneralPolicyPolicy: LinUCB with unique linear models
LinUCBHybridOptimizedPolicyPolicy: LinUCB with hybrid linear models
LinUCBHybridPolicyPolicy: LinUCB with hybrid linear models
mvrnormSimulate from a Multivariate Normal Distribution
OfflineBootstrappedReplayBanditBandit: Offline Bootstrapped Replay
OfflineDirectMethodBanditBandit: Offline Direct Methods
OfflineDoublyRobustBanditBandit: Offline Doubly Robust
OfflineLookupReplayEvaluatorBanditBandit: Offline Replay with lookup tables
OfflinePropensityWeightingBanditBandit: Offline Propensity Weighted Replay
OfflineReplayEvaluatorBanditBandit: Offline Replay
one_hotOne Hot Encoding of data.table columns
ones_in_zeroesA vector of zeroes and ones
OraclePolicyPolicy: Oracle
PlotPlot
plot.historyPlot Method for Contextual History
PolicyPolicy: Superclass
print.historyPrint Method for Contextual History
prob_winnerBinomial Win Probability
RandomPolicyPolicy: Random
sample_one_ofSample one element from vector or list
set_externalChange Default Graphing Device from RStudio
sherman_morrissonSherman-Morrisson inverse
sim_postBinomial Posterior Simulator
SimulatorSimulator
SoftmaxPolicyPolicy: Softmax
summary.historySummary Method for Contextual History
sum_ofSum of list
ThompsonSamplingPolicyPolicy: Thompson Sampling
UCB1PolicyPolicy: UCB1
UCB2PolicyPolicy: UCB2
value_remainingPotential Value Remaining
var_welfordWelford's variance
which_max_listGet maximum value in list
which_max_tiedGet maximum value randomly breaking ties
contextual documentation built on March 17, 2019, 5:05 p.m.