pegasos: pegasos

In wrathematics/ssvm: Simple SVM

Description

If intercept=TRUE, the data is not literally modified to include a column of ones. However, this is conceptually what we do.

Usage

pegasos(x, y, intercept = TRUE, k = 5, lambda = 0.1, niter = 1e+05)

Arguments

x	The features matrix.
y	The 2-class response vector. Note that the groups must be +/- 1. See recode()
intercept	Should the model include an intercept term?
k	TODO, currently ignored
lambda	Regularization parameter.
niter	The number of iterations.

Details

Pegasos SVM solver.

Value

An object of class pegasosSVM.

Author(s)

Drew Schmidt

References

Shalev-Shwartz, Shai, et al. "Pegasos: Primal estimated sub-gradient solver for svm." Mathematical programming 127.1 (2011): 3-30.

Examples

## Not run: 
library(ssvm)

y <- recode(iris$Species == "setosa")
x <- as.matrix(iris[, -5])

pegasos(x, y)

## End(Not run)

wrathematics/ssvm documentation built on May 4, 2019, 10:54 a.m.