classifier_accuracies: Classifier Accuracies

Description Usage Arguments

Description

Find the best classifier using leave-one-out cross validation (svm) and out-of-bag error (random forests). Returns a list of classifier results

Usage

1
classifier_accuracies(peaks, labels, training, min_peak_percentage)

Arguments

peaks

Boolean matrix of mass spectra rows with m/z columns, indicating if an m/z value corresponds to a peak.

labels

The correct classifications of the peaks.

training

The rows to actually use to train

minpeaks

How many "true" values must show up for a given m/z value for it to be considered a feature.


smanchan96/binspec documentation built on May 30, 2019, 3:06 a.m.