Cross-Validated Posterior Classification Summaries for a Naive Bayes model

Given a set of observed data including a categorical response variable y and a naiveBayes model of y, this function returns a cross validated confusion matrix by which to assess the model's posterior classification quality.

naive_classification_summary_cv(model, data, y, k = 10)

Arguments

model

a naiveBayes model object with categorical y

data

data frame including the variables in the model

y

a character string indicating the y variable in data

k

the number of folds to use for cross validation

Value

a list

Examples

data(penguins_bayes, package = "bayesrules")
example_model <- e1071::naiveBayes(species ~ bill_length_mm, data = penguins_bayes)
naive_classification_summary_cv(model = example_model, data = penguins_bayes, y = "species", k = 2)
#> $folds
#>   fold   Adelie  Chinstrap    Gentoo overall_accuracy
#> 1    1 0.961039 0.06451613 0.8750000        0.7674419
#> 2    2 0.960000 0.13513514 0.8666667        0.7500000
#> 
#> $cv
#>    species       Adelie  Chinstrap       Gentoo
#>     Adelie 96.05% (146)  0.00% (0)  3.95%   (6)
#>  Chinstrap  5.88%   (4) 10.29% (7) 83.82%  (57)
#>     Gentoo  7.26%   (9)  5.65% (7) 87.10% (108)
#>