Support Vector Machine Classification

Fits a radial basis Support Vector Machine Classification Model.

svmClassif(
  response = response,
  recipe = rec,
  folds = folds,
  train = train_df,
  test = test_df,
  gridNumber = 15,
  evalMetric = "bal_accuracy"
)

Arguments

response

The variable that is the response for analysis.

recipe

A recipe object.

folds

A rsample::vfolds_cv object.

train

The training data set.

test

Data frame/tibble. The testing data set.

gridNumber

Numeric. The size of the grid to tune on. Default is 15.

evalMetric

The classification metric you want to evaluate the model's accuracy on. Default is bal_accuracy. List of metrics available to choose from:

  • bal_accuracy

  • mn_log_loss

  • roc_auc

  • mcc

  • kap

  • sens

  • spec

  • precision

  • recall

Value

A list with the following outputs:

  • Training confusion matrix

  • Training model metric score

  • Testing confusion matrix

  • Testing model metric score

  • Final model chosen

  • Tuned model

Details

Note - tunes the following parameters:

  • cost: The cost of predicting a sample within or on the wrong side of the margin.

  • rbf_sigma: The precision parameter for the radial basis function.

Examples

library(easytidymodels)
library(dplyr)
library(recipes)
utils::data(penguins, package = "modeldata")
#Define your response variable and formula object here
resp <- "sex"
formula <- stats::as.formula(paste(resp, ".", sep="~"))
#Split data into training and testing sets
split <- trainTestSplit(penguins, stratifyOnResponse = TRUE,
responseVar = resp)
#Create recipe for feature engineering for dataset, varies based on data working with
rec <- recipe(formula, data = split$train) %>% step_knnimpute(!!resp) %>%
step_dummy(all_nominal(), -all_outcomes()) %>%
step_medianimpute(all_predictors()) %>% step_normalize(all_predictors()) %>%
step_dummy(all_nominal(), -all_outcomes()) %>% step_nzv(all_predictors()) %>%
step_corr(all_numeric(), -all_outcomes(), threshold = .8) %>% prep()
train_df <- bake(rec, split$train)
test_df <- bake(rec, split$test)
folds <- cvFolds(train_df)

#svm <- svmClassif(recipe = rec, response = resp, folds = folds,
#train = train_df, test = test_df)

#Confusion Matrix
#svm$trainConfMat

#Plot of confusion matrix
#svm$trainConfMatPlot

#Test Confusion Matrix
#svm$testConfMat

#Test Confusion Matrix Plot
#svm$testConfMatPlot