Xavier Llorà

Archive for 'Research' Category

« Previous Page — Next Page »

How big is big?

21 September 2006

Recently I have been working on a data-mining problem that requires supervised learning. The problem is not supposed to be big, just a few hundreds of features. The interesting issue is the number of records, around half a million or more. Most of the implementations of supervised learning algorithms available on the web are not designed with such a volume of data. Scalability of the algorithms becomes a clear issue when dealing such a volume of data. For instance, algorithms that scale as n^3 with respect to the data may become prohibitively costly for any feasible approach. Also, algorithms that require global processing, defeating efficient parallelization may not be an option either.

For such reasons, I started working last fall on efficient implementations of GBML algorithms focusing on (1) efficient implementations hacking the available hardware, (2) minimizing memory food prints required, and (3) massively exploiting the inherent parallelism of such methods. A few initial steps can be found here and here.

Evolving emotional prosody

17 September 2006

by Cecilia Ovesdotter Alm and Xavier Llorà (2006).
Proceedings of the Ninth International Conference on Spoken Language Processing (INTERSPEECH 2006), paper 1741. Also as IlliGAL TR No 2006018. Link to the PDF.

(more…)

Finally I am back up

17 September 2006

Vacations were over quite a while ago, but things kept piling on my desk. Finally I succeeded clearing some big chunks of matters out of it. Enough that I could go to the Design Theory and Methodology (DTM 2006) conference in Philadelphia last Tuesday. The conference is held as part of the ASME International Design Engineering Technical Conferences (IDETC 2006) . Yan Jin invited me to join the panel he organized on Intelligent Systems and Innovation. I had a great time joining the DTM people, very interesting work—if you are interested on innovation and creativity.

E2K: Evolution to knowledge

13 September 2006

by Xavier Llorà (2006).
ACM SIGEvolution, Volume 1 , Issue 3, pp. 10-17. Link to the Journal.
Also as IlliGAL TR No 2006022. Link to the PDF.

(more…)

GECCO 2006

8 July 2006

Well, it finally happened. GECCO started today at 8:30am. You can find live blogging at the IlliGAL Blogging.

Analyzing active interactive genetic algorithms using visual analytics

7 July 2006

by Xavier Llorà, Kumara Sastry , Francesc Alías, David E. Goldberg, and Michael Welge (2006).
Proceedings of the ACM Genetic and Evolutionary Computation Conference (GECCO 2006), pp. 1417–1418, ACM press. Also as IlliGAL TR No 2006004. Link to the PDF.

(more…)

Fast rule matching for Learning Classifier Systems via vector instructions

7 July 2006

by Xavier Llorà and Kumara Sastry (2006, accepted).
Proceedings of the ACM Genetic and Evolutionary Computation Conference (GECCO 2006), pp. 1513–1520, ACM press. Also as IlliGAL TR No 2006001. Link to the PDF.

(more…)

The χ-ary extended compact classifier system: Linkage learning in Pittsburgh LCS

7 July 2006

by Xavier Llorà, Kumara Sastry, David E. Goldberg, and Luis de la Ossa (2006).
To appear in the Proceedings of the International Workshop on Learning Classifier Systems (IWLCS 2006). Also as IlliGAL TR No 2006015. Link to the PDF.

(more…)

LCS and other GBML warming up for GECCO 2006

6 July 2006

The agenda for the Ninth International Workshop on Learning Classifier Systems (IWLCS’2006) can be found here. The workshop is coming with a list of very interesting papers and topics. We are looking forward to another edition of the workshop crowded of new and exiting ideas. If you are in GECCO, do not let it pass by ;).

By the way, if you want to dig a little further into the learning classifier systems (LCS) and other genetics-based machine learning (GBML) world, check the website of the NCSA/IlliGAL Gathering on Evolutionary Learning (NIGEL’2006). It contains the slides and videos of the talks on cutting-edge LCS and GBML research. It is just fresh out of the oven!

GALE is back!

10 May 2006

Yes, with the new update of the website GALE was unreachable for a while. Here it is the original code again. Please, take a few minutes to read this post. Hope you find it useful. Enjoy

GALE?

GALE (Genetic and Artificial Life Environment) is fine-grained parallel genetic algorithm for data mining. Its main contributions are simplicity and its knowledge-independent model. The simplicity of GALE relies in its fine-grained parallelism based on spreading the population (feasible solutions to the classification task) over a 2D grid. Thus, artificial evolution can be easily modeled in terms of neighborhood relations. These neighborhood relations define GALE as a massive parallel evolutionary model. On the other hand, GALE does not constrain the knowledge representation. It can evolve indistinctly rules, instances, partially defined instances, and decision trees (orthogonal, oblique, and multivariate based on nearest neighbor).

Where can I found more information about GALE?

GALE, as well as some results achieved using it, had been published on workshops, international conferences, and books, among others. I strongly suggest that if you are trying to use GALE for any purpose first read some of the papers that explain it. I would suggest that you check the proceeding of the Genetic and Evolutionary Computation Conference (GECCO) from 2001-2005, the books from the International Workshop on Learning Classifier Systems (IWLCS) from 2001-2005, and a two book chapter in Multi-Objective Machine Learning and Applications of Learning Classifier Systems.

Can I use GALE?

GALE is distributed under GNU GPL license for educational purposes. If this is not your case, please contact xllora@illigal.ge.uiuc.edu.

Download & documentation

The current release of GALE is version 0.9alpha. This release is written in Java, being some parts rewritten or removed, for educational purposes. Not all the current GALE knowledge representations and functionalities are included. As you see, it is an alpha, thus, it is possible that it still has some bugs that I have not fixed, yet. If you find some, please send them to xllora@illigal.ge.uiuc.edu. I will be happy to patch them.

The main source of documentation about GALE source code is in the reference manual. In the distribution file you will find the following stuff:

./README	->	A file with a little information.
./COPYING	->	The GNU GPL license.
./INSTALL	->	Some installation issues.
./gale_<version>.src.jar	->	The source code of GALE (.java files.)
./gale_<version>.bin.jar	->	The binaries of GALE (.class files.) You can run it using java -jar
./gale_<version>.doc.jar	->	The reference manual.
./data/	->	This directory contains some training data.
./cnf/	->	Some examples of configuration files are provided here.

Distribution files

Finally, here you can find GALE distribution files. Remember that you accept the terms of the GNU GPL license described above (you can find a copy inside the distribution file).

gale_0.9alpha.tar.bz2 (389Kb)

gale_0.9alpha.tar.gz (394Kb)

gale_0.9alpha.tar (675Kb)



« Previous Page — Next Page »