autoRLearn_: Advanced version of autoRLearn.
In DataSystemsGroupUT/SmartML: Machine Learning Automation

Description Usage Arguments Value

View source: R/autoRLearn_.R

Tunes the hyperparameters of the desired algorithm/s using either hyperband or BOHB.

autoRLearn_(
  df_train,
  df_test,
  maxTime = 10,
  models = c("randomForest", "naiveBayes", "boosting", "l2-linear-classifier", "svm"),
  optimizationAlgorithm = "hyperband",
  bw = 3,
  kde_type = "single",
  max_iter = 81,
  metric = "acc"
)

`df_train`	Dataframe of the training dataset. Assumes it is in perfect shape with all numeric variables and factor response variable named "class".
`df_test`	Dataframe of the test dataset. Assumes it is in perfect shape with all numeric variables and factor response variable named "class".
`maxTime`	Float representing the maximum time the algorithm should be run (in minutes).
`models`	List of strings denoting which algorithms to use for the process: "randomForest" - Random forests using the randomForest package "ranger - Random forests using the ranger package (unstable) "naiveBayes" - Naive bayes using the fastNaiveBayes package "boosting" - Gradient boosting using xgboost "l2-linear-classifier" - Linear primal Support vector machine from LibLinear "svm" - RBF kernel svm from e1071
`optimizationAlgorithm`	- String of which hyperparameter tuning algorithm to use: "hyperband" - Hyperband with uniformly initiated parameters "bohb" - Hyperband with bayesian optimization as described on F. Hutter et al 2018 paper BOHB. Has extra parameters bw and kde_type
`bw`	- (only applies to BOHB) Double representing how much should the KDE bandwidth be widened. Higher values allow the algorithm to explore more hyperparameter combinations
`kde_type`	- (only applies to BOHB) String representing whether a model's hyperparameters should be tuned individually of each other or have their probability densities multiplied: "single" - each hyperparameter has its own expected improvement calculated "mixed" - all hyperparameters' probabilty densities are multiplied and only one mixed expected improvement is calculated
`max_iter`	- (affects both hyperband and BOHB) Integer representing the maximum number of iterations that one successive halving run can have
`metric`	String of the evaluation metric to be used in the model performance optimization: "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.

List of Results

perf - Evaluated metric of the best performing model on the test data
pred - prediction on the test data using the best model
model - best model object
best_models - table with the best hyperparameters found for the selected models.