runClassifier: Fit a classifier model.
In DataSystemsGroupUT/SmartML: Machine Learning Automation

Description Usage Arguments Value Examples

View source: R/runClassifier.R

Run the classifier on a training set and measure performance on a validation set.

runClassifier(
  trainingSet,
  validationSet,
  params,
  classifierAlgorithm,
  metric = "acc",
  interp = 0
)

`trainingSet`	Dataframe of the training set.
`validationSet`	Dataframe of the validation Set.
`params`	A string character of parameter configuration values for the current classifier to be tuned (parameters are separated by #) and can be obtained from `params` out of resulted list after running `autoRLearn` function.
`classifierAlgorithm`	String character of the name of classifier algorithm used now. "svm" - Support Vector Machines from e1071 package, "naiveBayes" - naiveBayes from e1071 package, "randomForest" - randomForest from randomForest package, "lmt" - LMT Weka classifier trees from RWeka package, "lda" - Linear Discriminant Analysis from MASS package, "j48" - J48 Weka classifier Trees from RWeka package, "bagging" - Bagging Classfier from ipred package, "knn" - K nearest Neighbors from FNN package, "nnet" - Simple neural net from nnet package, "C50" - C50 decision tree from C5.0 pacakge, "rpart" - rpart decision tree from rpart package, "rda" - regularized discriminant analysis from klaR package, "plsda" - Partial Least Squares And Sparse Partial Least Squares Discriminant Analysis from caret package, "glm" - Fitting Generalized Linear Models from stats package, "deepboost" - deep boost classifier from deepboost package.
`metric`	Metric string character to be used in evaluation: "acc" - Accuracy, "avg-fscore" - Average of F-Score of each label, "avg-recall" - Average of Recall of each label, "avg-precision" - Average of Precision of each label, "fscore" - Micro-Average of F-Score of each label, "recall" - Micro-Average of Recall of each label, "precision" - Micro-Average of Precision of each label
`interp`	Boolean representing if interpretability is required or not (Default = 0).

List of performance on validationSet named perf, model fitted on trainingSet named m, predictions on test set pred, and interpretability plots named interpret in case of interp = 1

## Not run: 
result1 <- autoRLearn(10, 'sampleDatasets/shuttle/train.arff', 'sampleDatasets/shuttle/test.arff')
dataset <- datasetReader('/Datasets/irisTrain.csv', '/Datasets/irisTest.csv')
result2 <- runClassifier(dataset$Train, dataset$Test, result1$params, result1$clfs)

## End(Not run)