prediction.metrics: Prediction Metric Calculations

In OmicsMarkeR: Classification and Feature Selection for 'Omics' Datasets

Description

Performance evaluation of all fitted models. This function concisely provides model performance metrics, including confusion matrix and ROC.

Usage

prediction.metrics(finalModel, method, raw.data, inTrain, outTrain, features,
  bestTune, grp.levs, stability.metric)

Arguments

finalModel	List of fitted models
method	Vector of strings dictating the models that were fit
raw.data	Original dataset prior to any training subset
inTrain	List of training indicies for each feature selection run
outTrain	List of testing data indicies for each feature selection run
features	List of selected features for each model
bestTune	List of parameters that have been optimized for the each respective model
grp.levs	Vector of group levels
stability.metric	A character object specifying the stability metric

Value

Returns a dataframe consisting of each feature selection runs evaluated Accuracy, Kappa, ROC.AUC, Sensitivity, Specificity, Positive Predictive Value, and Negative Predictive Value.

See Also

performance.stats, perf.calc caret function confusionMatrix

OmicsMarkeR documentation built on April 28, 2020, 6:54 p.m.