mClass: m- Class classifier

In MarcoDVisser/rML: Machine Learning routines for R

Description

Trains a m-class classifer from dataset trainingdata

uses a trained classifier to predict classes

uses a trained classifier to predict classes and displayes random images with predictions on screeen with a delay

Usage

mClass(trainingdata, y, lambda = 1)

mPred(Data, Theta)

digitAnim(Data, Theta, y, maxit = 10)

Arguments

trainingdata	numeric matrix with training data
y	classes corresponding to the rows of trainingdata
lambda	regularization coefficient
Data	numeric matrix with data to classify
Theta	a previously trained classifier
Data	numeric matrix with data to classify
Theta	a previously trained classifier
maxit	maximum number of random displayes
y	True classes used in training

Author(s)

Marco D. Visser

Marco D. Visser

Marco D. Visser

Examples

data(digits)
X<-digits[[1]]
y<-digits[[2]]
trained<-mClass(X,y)
data(digits)
X<-digits[[1]]
y<-digits[[2]]
trained<-mClass(X,y)
pred<-mPred(X,trained)
image1200<-t(as.array(X[1200,]))
pred1200<-mPred(image1200,trained)
## force correct rotation of matrix image
f <- function(m) t(m)[,nrow(m):1]
image(f(matrix(X[1200,]),ncol=20,nrow=20),main=pred1200)
data(digits)
X<-digits[[1]]
y<-digits[[2]]
trained<-mClass(X,y)

MarcoDVisser/rML documentation built on May 7, 2019, 2:49 p.m.