trainAlgorithm: Train a learning algorithm and return the resulting model.

In AleMorales/SeedSorter: Supervised classification of seeds from flow cytometry

Description

Train a learning algorithm and return the resulting model.

Usage

trainAlgorithm(algorithm = "xgboost", task, osw.rate = 10)

Arguments

algorithm	Name of the algorithm to be used.
task	A classification task as returned by createTrainingTask().

Details

The list of algorithms that can be used are:

• xgboost: The XGBoost algorithm.

• knn: The fast k-nearest neighbour algorithm fromt the FNN package.

• lda: Linear discriminant analysis.

• qda: Quadratic discriminant analysis applied to principal components.

• naiveBayes: Naive Bayes classifier.

• logistic: Logistic regression regularized with elasticnet grid penalty.

• svm: Support vector machine with radial basis kernel (aka Gaussian kernel).

• randomForest: Multithreaded random forest as implemented in the package randomForestSRC.

• kmeans: k-means clustering. This method will cluster the data into two clusters and assign the minority cluster to seed or waste depending on the parameter seed_minority.

Value

A trained model that can be used to make predictions.

AleMorales/SeedSorter documentation built on Feb. 12, 2020, 4:13 a.m.