Skip to content
Matteo Santoro edited this page Jul 1, 2013 · 28 revisions

Table of Contents

Description

GURLS – which stands for Grand Unified Regularized Least Squares – is a software library for regression and (multiclass) classification based on the Regularized Least Squares (RLS) loss function. GURLS takes advantage of some favorable properties of regularized least squares algorithm and is tailored to deal in particular with multi-category/multi-label problems. The library comprises four main modules. GURLS and bGURLS – both implemented in Matlab – are aimed at solving learning problems with small/medium and large-scale datasets respectively. GURLS++ and bGURLS++ are their C++ counterparts. In terms of memory, we consider to be big those data that cannot fully reside in RAM without any memory mapping techniques – such as swapping. Data that can fully reside in RAM are considered to be small/medium. Learning with small/medium data can be carried out via GURLS or GURLS++. Learning with big data can be carried out with the dedicated modules bGURLS or bGURLS++, though limited to linear models, and provided that it is possible to store a dxd matrix in memory, where d is the number of features. Due to programming language constraints, there are some differences between bGURLS and bGURLS++. In the Figure below we depicted the different scenarios where the different packages can be applied and show how the two modules are able to deal with big learning scenarios by means of parallel or serial computing.

Gurls ScenariosGurls Scenarios

License

GURLS is distributed under the BSD license. This means that it is free for both academic and commercial use. If you are going to use GURLS in your scientific work, please cite the toolbox, the main website and the paper:

Tacchetti, A., P. Mallapragada, M. Santoro, and L. Rosasco; GURLS: a Toolbox for Large Scale Multiclass Learning, presented at Workshop: "Big Learning: Algorithms, Systems, and Tools for Learning at Scale" at NIPS 2011, December 16-17 2011, Sierra Nevada, Spain.

Download

Download the source from the Git repository http://github.com/CBCL/GURLS

Installation instructions

For GURLS installation instruction go to page Installing GURLS For bGURLS installation instruction go to page Installing bGURLS For GURSL++ and bGURLS++ installation instruction go to page Installing GURSL++/bGURLS++

Usage

Design

To learn about GURLS design go to page GURLS Design.

Hello World

For GURLS helloworld go to page Hello World in GURLS.

For GURLS++ helloworld go to page Hello World in GURLS++

User Manuals

For learning how to use GURLS go to page GURLS Usage.

For learning how to use GURLS++ go to page GURLS++ Usage.

For learning how to use bGURLS go to page bGURLS Usage.

For learning how to use bGURLS go to page bGURLS++ Usage.

A comprehensive user and developer manual can be found here here.