dummy-link

RegERMs

Regularised Empirical Risk Minimisation Framework (SVMs, LogReg, Linear Regression) in Julia

Readme

RegERMs Logo

RegERMs.jl

Build Status Coverage Status RegERMs RegERMs

This package implements several machine learning algorithms in a regularised empirical risk minimisation framework (SVMs, LogReg, Linear Regression) in Julia.

Quick start

Some examples:

using RegERMs

# define some toy data (XOR - example)
np = 100
nn = 100
X = [randn(int(np/2),1)+1 randn(int(np/2),1)+1; randn(int(np/2-0.5),1)-1 randn(int(np/2-0.5),1)-1;
     randn(int(nn/2),1)+1 randn(int(nn/2),1)-1; randn(int(nn/2-0.5),1)-1 randn(int(nn/2-0.5),1)+1] # examples with 2 features
y = int(vec([ones(np,1); -ones(nn,1)]))       # binary class values

# use rbf kernel by using mercer map
map = MercerMap(X, :rbf)
X = RegERMs.apply(map)

# choose (linear) SVM as learning algorithm with regularization parameter 0.1
svm = SVM(X, y; λ=0.1)

# get a solution 
model = optimize(svm)

# make predictions and compute accuracy
ybar = predict(model, X)
acc = mean(ybar .== y)

Documentation

Full documentation available at Read the Docs.

First Commit

06/13/2014

Last Touched

6 months ago

Commits

46 commits

Used By: