Contextual: Multi-Armed Bandits in R

Overview

R package facilitating the simulation and evaluation of context-free and contextual Multi-Armed Bandit policies.

The package has been developed to:

• Ease the implementation, evaluation and dissemination of both existing and new contextual Multi-Armed Bandit policies.
• Introduce a wider audience to contextual bandit policies' advanced sequential decision strategies.

Package links:

• Package website
• Package development FAQ

Installation

To install contextual from CRAN:

install.packages('contextual')

To install the development version (requires the devtools package):

install.packages("devtools")
devtools::install_github('Nth-iteration-labs/contextual')

When working on or extending the package, clone its GitHub repository, then do:

install.packages("devtools")
devtools::install_deps(dependencies = TRUE)
devtools::build()
devtools::reload()

clean and rebuild...

Overview of core classes

Contextual consists of six core classes. Of these, the Bandit and Policy classes are subclassed and extended when implementing custom (synthetic or offline) bandits and policies. The other four classes (Agent, Simulator, History, and Plot) are the workhorses of the package, and generally need not be adapted or subclassed.

Documentation

See the demo directory for practical examples and replications of both synthetic and offline (contextual) bandit policy evaluations.

When seeking to extend contextual, it may also be of use to review "Extending Contextual: Frequently Asked Questions", before diving into the source code.

How to replicate figures from two introductory context-free Multi-Armed Bandits texts:

• Replication of figures from Sutton and Barto, "Reinforcement Learning: An Introduction", Chapter 2
• Replication of figures from "Bandit algorithms for website optimization" by John Miles White

Basic, context-free multi-armed bandit examples:

• Basic MAB Epsilon Greedy evaluation
• Synthetic MAB policy comparison
• Replication Eckles & Kaptein (Bootstrap Thompson Sampling)

Examples of both synthetic and offline contextual multi-armed bandit evaluations:

• Synthetic cMAB policy comparison

• Offline Evaluation Data Set - Basic Replay Bandit: Ratings

• Offline Evaluation Data Set - Bootstrapped Replay Bandit: Carskit DePaul Movies

• Offline Evaluation Data Set - Lookup Table Replay Bandit: MovieLens 10M

• Offline Bandits and Simpson's Paradox

An example how to make use of the optional theta log to create interactive context-free bandit animations:

• Interactive, animated versions of Epsilon Greedy, UCB1 and Thompson Sampling policies

Some more extensive vignettes to get you started with the package:

• Getting started: running simulations
• Offline evaluation: replication of Li et al (2010)
• Class reference

Paper offering a general overview of the package's structure & API:

• An introduction to multi-armed bandit problems and the use of the R package contextual paper.

Policies and Bandits

Overview of contextual's growing library of contextual and context-free bandit policies:

General	Context-free	Contextual	Other
Random Oracle Fixed	Epsilon-Greedy Epsilon-First UCB1, UCB2 Thompson Sampling BootstrapTS Softmax Gradient Gittins	CMAB Naive Epsilon-Greedy Epoch-Greedy LinUCB (General, Disjoint, Hybrid) Linear Thompson Sampling ProbitTS LogitBTS GLMUCB	Lock-in Feedback (LiF)

Overview of contextual's bandit library:

Basic Synthetic	Contextual Synthetic	Offline	Continuous
Basic Bernoulli Bandit Basic Gaussian Bandit	Contextual Bernoulli Contextual Logit Contextual Hybrid Contextual Linear Contextual Wheel	Replay Evaluator Bootstrap Replay Propensity Weighting Direct Method Doubly Robust	Continuum

Alternative parallel backends

By default, "contextual" uses R's built-in parallel package to facilitate parallel evaluation of multiple agents over repeated simulation. See the demo/alternative_parallel_backends directory for several alternative parallel backends:

• Microsoft Azure VM's using doAzureParallel.
• Redis using doRedis.
• MPI (Message-Passing Interface) using Rmpi and doMPI.

Maintainers

Robin van Emden: author, maintainer* Maurits Kaptein: supervisor*

* Tilburg University / Jheronimus Academy of Data Science.

If you encounter a clear bug, please file a minimal reproducible example on GitHub.