R package arulesCBA: Classification Based on Association Rules

CRAN version CRAN RStudio mirror downloads R-CMD-check CRAN RStudio mirror downloads

The R package arulesCBA (Hahsler et al, 2020) is an extension of the package arules to perform association rule-based classification. The package provides the infrastructure for class association rules and implements associative classifiers based on the following algorithms:

The package also provides the infrastructure for associative classification (supervised discetization, mining class association rules (CARs)), and implements various association rule-based classification strategies (first match, majority voting, weighted voting, etc.).

Installation

Stable CRAN version: install from within R with

install.packages("arulesCBA")

Current development version:

devtools::install_github("ianjjohnson/arulesCBA")

Usage

library("arulesCBA")
data("iris")

Learn a classifier.

classifier <- CBA(Species ~ ., data = iris)
classifier
## CBA Classifier Object
## Formula: Species ~ .
## Number of rules: 6
## Default Class: NA
## Classification method: first  
## Description: CBA algorithm (Liu et al., 1998)

Inspect the rulebase.

inspect(rules(classifier), linebreak = TRUE)
##     lhs                            rhs                  support confidence coverage lift count size coveredTransactions totalErrors
## [1] {Petal.Length=[-Inf,2.45)}  => {Species=setosa}        0.33       1.00     0.33  3.0    50    2                  50          50
## [2] {Sepal.Length=[6.15, Inf],                                                                                                     
##      Petal.Width=[1.75, Inf]}   => {Species=virginica}     0.25       1.00     0.25  3.0    37    3                  37          13
## [3] {Sepal.Length=[5.55,6.15),                                                                                                     
##      Petal.Length=[2.45,4.75)}  => {Species=versicolor}    0.14       1.00     0.14  3.0    21    3                  21          13
## [4] {Sepal.Width=[-Inf,2.95),                                                                                                      
##      Petal.Width=[1.75, Inf]}   => {Species=virginica}     0.11       1.00     0.11  3.0    17    3                   5           8
## [5] {Petal.Width=[1.75, Inf]}   => {Species=virginica}     0.30       0.98     0.31  2.9    45    2                   4           6
## [6] {}                          => {Species=versicolor}    0.33       0.33     1.00  1.0   150    1                  33           6

Make predictions for the first few instances of iris.

predict(classifier, head(iris))
## [1] setosa setosa setosa setosa setosa setosa
## Levels: setosa versicolor virginica

References