FuzzyClass

# Installation
install.packages("devtools")
devtools::install_github("Jodavid/FuzzyClass")

# package import
library(FuzzyClass)

library(FuzzyClass)
library(caret)
#> Carregando pacotes exigidos: ggplot2
#> Carregando pacotes exigidos: lattice
#> Warning in system("timedatectl", intern = TRUE): execução do comando
#> 'timedatectl' teve status 1

#' ---------------------------------------------
#' The following shows how the functions are used:
#' --------------
#' Reading a database:
#'
#' Actual training data:
data(VirtualRealityData)

VirtualRealityData <- as.data.frame(VirtualRealityData)

# Splitting into Training and Testing
split <- caTools::sample.split(t(VirtualRealityData[,1]), SplitRatio = 0.7)
Train <- subset(VirtualRealityData, split == "TRUE")
Test <- subset(VirtualRealityData, split == "FALSE")
# ----------------

test = Test[,-4]

# --------------------------------------------------
# Fuzzy Gaussian Naive Bayes with Fuzzy Parameters


fit_FGNB <- GauNBFuzzyParam(train =  Train[,-4],
                                    cl = Train[,4], metd = 1, cores = 1)


print(fit_FGNB)
#> 
#> Fuzzy Gaussian Naive Bayes Classifier for Discrete Predictors
#> 
#> Variables:
#> [1] "V1" "V2" "V3"
#> Class:
#> [1] "1" "2" "3"
saida <- predict(fit_FGNB, test)
confusionMatrix(factor(Test[,4]), saida)
#> Confusion Matrix and Statistics
#> 
#>           Reference
#> Prediction  1  2  3
#>          1 53  4  0
#>          2  7 48 10
#>          3  1 12 45
#> 
#> Overall Statistics
#>                                           
#>                Accuracy : 0.8111          
#>                  95% CI : (0.7462, 0.8655)
#>     No Information Rate : 0.3556          
#>     P-Value [Acc > NIR] : <2e-16          
#>                                           
#>                   Kappa : 0.7163          
#>                                           
#>  Mcnemar's Test P-Value : 0.5724          
#> 
#> Statistics by Class:
#> 
#>                      Class: 1 Class: 2 Class: 3
#> Sensitivity            0.8689   0.7500   0.8182
#> Specificity            0.9664   0.8534   0.8960
#> Pos Pred Value         0.9298   0.7385   0.7759
#> Neg Pred Value         0.9350   0.8609   0.9180
#> Prevalence             0.3389   0.3556   0.3056
#> Detection Rate         0.2944   0.2667   0.2500
#> Detection Prevalence   0.3167   0.3611   0.3222
#> Balanced Accuracy      0.9176   0.8017   0.8571

saida <- predict(fit_FGNB, test, type = "matrix")