estimateRelativeOverfitting
Estimate relative overfitting.
Result of multi-criteria tuning.
Internal construction / wrapping of learner object.
Generate dummy variables for factor features.
Filter features by thresholding filter values.
Convert a machine learning benchmark / demo object from package mlbench to a task.
Estimate the residual variance
Iris cost-sensitive classification task
Return measures used in benchmark.
Result of tuning.
Extract the feature selection results from a benchmark result.
BenchmarkResult object.
Return task ids used in benchmark.
Return error message of FailureModel.
Get feature names of task.
Get tuning parameters from a learner of the caret R-package.
European Union Agricultural Workforces clustering task
Get number of observations in task.
Extract the aggregated performance values from a benchmark result.
Aggregation methods.
Deprecated, use getPredictionProbabilities
instead.
Get the tuned hyperparameter settings from a nested tuning.
Get the class weight parameter of a learner.
Get the class levels for classification and multilabel tasks.
Returns the filtered features.
getHomogeneousEnsembleModels
Deprecated, use getLearnerModel
instead.
crossover
Create a custom imputation method.
Returns a list of mlr's options
Returns the selected feature set and optimization path after training.
Specifiy your own aggregation of measures
Create control structures for tuning.
generatePartialPredictionData
Generate partial predictions
Generate classifier calibration data.
Get the id of the task.
Get current parameter settings for a learner.
Predict new data.
Fuse learner with the bagging technique.
Extract the tuning results from a benchmark result.
Extract costs in task.
Confusion matrix.
Iris classification task
Create control structures for multi-criteria tuning.
Impute and re-impute data
Is the model a FailureModel?
Plot filter values using ggvis.
Find matching measures.
Summarizes factors of a data.frame by tabling them.
Induced model of learner.
NCCTG Lung Cancer survival task
Wraps a regression learner for use in cost-sensitive learning.
Fuse learner with simple downsampling (subsampling).
Perform overall Friedman test for a BenchmarkResult.
Get underlying R model of learner integrated into mlr.
Description object for task.
ResampleResult object.
Get predictions from resample results.
Boston Housing regression task
Failure model.
Get default measure.
Generate binary classification predictions via ROCR ROC curves.
Aggregation object.
Fuse learner with a feature filter method.
Drop some features of task.
Return learners used in benchmark.
Get the name(s) of the target column(s).
Show and visualize the steps of feature selection.
getNestedTuneResultsOptPathDf
Get the opt.path
s from each tuning step from the outer resampling.
Convert large/infinite numeric values in a data.frame or task.
Performance measures.
Find matching learning algorithms.
Return measures IDs used in benchmark.
Result of feature selection.
Create a description object for a resampling strategy.
Calculates feature filter values.
Return learner ids used in benchmark.
Get response / truth from prediction object.
getPredictionProbabilities
Get probabilities for some classes.
Fuse learner with multiclass method.
Set the probability threshold the learner should use.
Generate a fixed holdout instance for resampling.
Sonar classification task
Downsample (subsample) a task or a data.frame.
Subset data in task.
Yeast multilabel classification task
Create a feature filter
Get a description of all possible parameter settings for a learner.
Wisconsin Breast Cancer classification task
Built in imputation methods
The built-ins are:
imputeConstant(const)
for imputation using a constant value,imputeMedian()
for imputation using the median,imputeMode()
for imputation using the mode,imputeMin(multiplier)
for imputing constant values shifted below the minimum
usingmin(x) - multiplier * diff(range(x))
,imputeMax(multiplier)
for imputing constant values shifted above the maximum
usingmax(x) + multiplier * diff(range(x))
,imputeNormal(mean, sd)
for imputation using normally
distributed random values. Mean and standard deviation will be calculated
from the data if not provided.imputeHist(breaks, use.mids)
for imputation using random values
with probabilities calculated usingtable
orhist
.imputeLearner(learner, preimpute)
for imputations using the response
of a classification or regression learner. Convert BenchmarkResult to a rank-matrix.
Creates a measure for non-standard misclassification costs.
Configures the behavior of the package.
Visualizes a learning algorithm on a 1D or 2D data set.
List of supported learning algorithms.
Remove constant features from a data set.
PimaIndiansDiabetes classification task
Motor Trend Car Road Tests clustering task
Merges small levels of factors into new level.
Plots results from generateROCRCurvesData using ggvis.
generateRankMatrixAsBarData
Generate data to plot a rank-matrix as a barplot.
makeMultilabelBinaryRelevanceWrapper
Use binary relevance method to create a multilabel learner.
Create a stacked learner object.
Fuse learner with the bagging technique and oversampling for imbalancy correction.
Plots results from generateROCRCurvesData using ggplot2.
Remove hyperparameters settings of a learner.
Perform a posthoc Friedman-Nemenyi test.
Returns the optimal hyperparameters and optimization path after training.
Converts predictions to a format package ROCR can handle.
Instantiates a resampling strategy object.
Plot filter values using ggplot2.
Train a learning algorithm.
Plot a partial prediction with ggplot2
Prediction from resampling.
Create a Trellis-plot for a selected measure.
Fuse learner with an imputation method.
Generate threshold vs. performance(s) for 2-class classification.
Set threshold of prediction object.
Plot calibration data using ggplot2.
Convert arguments to control structure.
Synthetic Minority Oversampling Technique to handle class imbalancy in binary classification.
generateBenchmarkSummaryData
Generate data for a benchmark-summary plot.
generateCritDifferencesData
Generate data for critical-differences plot.
Create learner object.
Get formula of a task.
Merge different tasks of BenchmarkResult objects
Fuse learner with feature selection.
Plots a ROC curve using ggplot2
mergeBenchmarkResultLearner
Merge different learners of BenchmarkResult objects
getMultilabelBinaryPerformances
Retrieve binary classification measures for multilabel classification predictions.
generateLearningCurveData
Generates a learning curve
Prediction object.
plotPartialPredictionGGVIS
Plot a partial prediction using ggvis
Get a summarizing task description.
Plot learning curve data using ggvis.
Fuse learner with SMOTE oversampling for imbalancy correction in binary classification.
Create a classification, regression, survival, cluster, cost-sensitive classification or
multilabel task.
Calculates feature filter values.
Get number of features in task.
Summarize columns of data.frame or task.
Hyperparameter tuning for multiple measures at once.
makeWeightedClassesWrapper
Wraps a classifier for weighted fitting where each class receives a weight.
makeCustomResampledMeasure
Construct your own resampled performance measure.
Create control structures for feature selection.
Extract the test performance values from a benchmark result.
makeModelMultiplexerParamSet
Creates a parameter set for model multiplexer tuning.
Plot threshold vs. performance(s) for 2-class classification using ggplot2.
Get the type of the task.
Tune prediction threshold.
Feature selection by wrapper approach.
Fuse learner with preprocessing.
Fuse learner with preprocessing
Plot threshold vs. performance(s) for 2-class classification using ggvis.
Get target data of task.
Re-impute a data set
Predict new data with an R learner.
Wisonsin Prognostic Breast Cancer (WPBC) survival task
Over- or undersample binary classification task to handle class imbalancy.
Set the hyperparameters of a learner object.
Normalize features
Benchmark experiment for multiple learners and tasks.
Set, add, remove or query properties of learners
Join some class existing levels to new, larger class levels for classification problems.
Plot critical differences for a selected measure.
makeCostSensClassifWrapper
Wraps a classification learner for use in cost-sensitive learning.
Deprecated, use hasLearnerProperties
instead.
Set the id of a learner object.
Train an R learner.
plotTuneMultiCritResultGGVIS
Plots multi-criteria results after tuning using ggvis.
Plot a rank-matrix as a barplot.
Fuse learner with simple ove/underrsampling for imbalancy correction in binary classification.
Plot a benchmark-summary.
Fuse learner with tuning.
Fit models according to a resampling strategy.
List filter methods
getStackedBaseLearnerPredictions
Returns the predictions for each base learner.
Hyperparameter tuning.
Set the type of predictions the learner should return.
Extract the predictions from a benchmark result.
Plot learning curve data using ggplot2.
Measure performance of prediction.
Set aggregation function of measure.
Visualize binary classification predictions via ViperCharts system.
Create model multiplexer for model selection to tune over multiple possible models.
Extract data in task.
Construct performance measure.
Extract the feature selection results from a benchmark result.
Plots multi-criteria results after tuning using ggplot2.
Only exported for internal use.
makeCostSensWeightedPairsWrapper
Wraps a classifier for cost-sensitive learning to produce a weighted pairs model.