classifast: Compare different classification methods on multivariate data

In Richyrecarey/classifast: Automatic Testing And Stacking Of Classifiers For Predicting

Description

Compare different classification methods on multivariate data

Usage

classifast(x, y, prob = 0.65, method = c("simple"),
  kfold = ifelse(nrow(x) < 100, floor(nrow(x)/15), 10), cv.iter = 1,
  timing = FALSE)

Arguments

x	Data frame or matrix with multivariate data with n observations (rows) and p variables (cols)
y	A factor with the labels of the rows of x
prob	Percentage p for the split train-test data. (1-prob)% is used for testing.
method	Vector of the methods wanted. By default, "simple" gives you various lineal classifiers. Other possibilities are: - "log": Logistic or multinomial linear logistic regression via neural networks - "svm": Support Vector Machines with Radial Kernel - "knn": kNN with cross-validation choosing of K - "rforest": Random Forest - "simple": Trains "log", "svm", "knn" and "rforest". - "all": All implemented classifiers (time consuming)
kfold	Number of folds in the cross validation estimation
cv.iter	Number of iterations to do with cross validation.
timing	if TRUE, shows you prediction of executing time. Feel free to ask the models we use.

Value

Not yet

Examples

Not yet

Richyrecarey/classifast documentation built on May 23, 2019, 1:09 p.m.