pcd_classifier: The PCD classifier based on the cover of the two (or more)...

In Artur-man/PCDSL: Proximity Catch Digraphs: Statistical Learning

The PCD classifier based on the cover of the two (or more) classes wherein the target class points inside the convex hull are classified with n-simplices, and the ones that are outside with outer simplices.

Description

The PCD classifier based on the cover of the two (or more) classes wherein the target class points inside the convex hull are classified with n-simplices, and the ones that are outside with outer simplices.

Usage

pcd_classifier(data, classes, map = "pe", p_pcd = 1)

Arguments

data	An n-by-d matrix of the training data set.
classes	A vector of length n indicating the labels of the classes.
map	The Proximity Map associated with the classifier. "pe" for the Proportional-edge proximity maps and "cs" for the Central-similarity proximity maps.
p_pcd	The value of the parameter associated with the proximity map.

Value

A proximity catch digraph (PCD).

Examples

# input parameters
ntest <- 100     # test data size for each class
nx <- 300        # training data size of x class (majority)
r <- 0.1         # Imbalance Ratio
de <- 0.5        # delta, the overlapping parameter
dimx <- 2        # number of dimensions

# training the classifier
set.seed(1)
x0 <- matrix(runif(dimx*nx,0,1),nrow=nx)
x1 <- matrix(runif(dimx*nx*r,de,1+de),nrow=nx*r)
x <- rbind(x0,x1)
classes <- rep(1:2,c(nx,nx*r))
graph_pcd <- pcd_classifier(x,classes,map="pe",p_pcd=1)

Artur-man/PCDSL documentation built on Feb. 24, 2024, 11:15 p.m.