Xavier Llorà » GBML & LCS

LCS and Software Development

Xavier — Fri, 18 Jun 2010 16:43:25 +0000

“On the Road to Competence” is a slide deck by Jurgen Appelo with interesting analogies between learning classifier systems and software development. Definitely worth taking a look at it.

Large Scale Data Mining using Genetics-Based Machine Learning

Xavier — Wed, 15 Jul 2009 21:56:17 +0000

Below you may find the slides of the GECCO 2009 tutorial that Jaume Bacardit and I put together. Hope you enjoy it.

Slides

Abstract

We are living in the peta-byte era.We have larger and larger data to analyze, process and transform into useful answers for the domain experts. Robust data mining tools, able to cope with petascale volumes and/or high dimensionality producing human-understandable solutions are key on several domain areas. Genetics-based machine learning (GBML) techniques are perfect candidates for this task, among others, due to the recent advances in representations, learning paradigms, and theoretical modeling. If evolutionary learning techniques aspire to be a relevant player in this context, they need to have the capacity of processing these vast amounts of data and they need to process this data within reasonable time. Moreover, massive computation cycles are getting cheaper and cheaper every day, allowing researchers to have access to unprecedented parallelization degrees. Several topics are interlaced in these two requirements: (1) having the proper learning paradigms and knowledge representations, (2) understanding them and knowing when are they suitable for the problem at hand, (3) using efficiency enhancement techniques, and (4) transforming and visualizing the produced solutions to give back as much insight as possible to the domain experts are few of them.

This tutorial will try to answer this question, following a roadmap that starts with the questions of what large means, and why large is a challenge for GBML methods. Afterwards, we will discuss different facets in which we can overcome this challenge: Efficiency enhancement techniques, representations able to cope with large dimensionality spaces, scalability of learning paradigms. We will also review a topic interlaced with all of them: how can we model the scalability of the components of our GBML systems to better engineer them to get the best performance out of them for large datasets. The roadmap continues with examples of real applications of GBML systems and finishes with an analysis of further directions.

NIGEL 2006 Part VI: Bacardit

Xavier — Mon, 13 Jul 2009 23:54:27 +0000

After coming back from GECCO I just uploaded the last of the NIGEL 2006 talks at LCS & GBML Central. This last talk was by Jaume Bacardit and GBML for protein structure prediction.

NIGEL 2006 Part V: Bernardó vs. Lanzi

Xavier — Mon, 29 Jun 2009 18:30:14 +0000

After the vacation break, two more NIGEL 2006 talks are available at LCS & GBML Central. This week Ester Bernardó presents how LCS can perform in the presence of class imbalance, whereas Lanzi continues his quest on computed predictions.

NIGEL 2006 Part IV: Llorà vs. Casillas

Xavier — Mon, 08 Jun 2009 21:20:02 +0000

Two more NIGEL 2006 talks are available at LCS & GBML Central. This week Xavier Llorà presents how linkage learning can be achieve in Pittsburgh LCS, whereas Jorge Casillas reviews his work using XCS and Fuzzy LCS.

NIGEL 2006 Part III: Butz vs. Barry

Xavier — Tue, 19 May 2009 14:38:05 +0000

NIGEL 2006 talks is available at LCS & GBML Central. This week Martin Butz review reviews the state of the union of XCS, where as Alwyn Barry introduces the theoretical framework for LCS that he and Jan Drugowitsch worked on.

NIGEL 2006 Part II: Dasgupta vs. Booker

Xavier — Mon, 11 May 2009 16:08:38 +0000

The second weekly release of NIGEL 2006 talks is available at LCS & GBML Central. This week Dipankar Dasgupta reviews the negative selection algorithm, where as Lashon Booker travels in time to the past and future of learning classifier systems.

Transcoding NIGEL 2006 videos

Xavier — Mon, 04 May 2009 20:13:23 +0000

Last week Pier Luca Lanzi was visiting IlliGAL. Yesterday, before he left for Chicago, we went for one last brunch. He mentioned that he liked a lot the videos we shot during NIGEL 2006. Thinking about it we agreed would be useful to recover the videos and upload them into some of the usual video sharing site suspects. Currently they are hosted, for long term storage purposes, at NCSA’s web archive. I spent sometime retrieving them from the archive (they are pretty fat and encoded in wmv) and I stated transcoding it in m4a. My plan? Make them available via Vimeo and LCS & GBML Central. Also, I will be uploading the presentation slides to SlideShare and also make them available via LCS & GBML Central.

Update: The first two videos (Wilson and Goldberg) are already available at LCS & GBML Central.

GECCO 2007 Hat Trick

Xavier — Fri, 13 Jul 2007 03:46:07 +0000

Yup, never happened to me before. In John Koza’s words: Hat Trick, best paper in EDA track and two Bronze medals in the HUMIES competition. I just think that I was lucky and blessed to have the pleasure to work with such great people. It would not have happened if it wasn’t for them. Below you can find some of the presentations for the above Hat Trick, as well as some other work I presented on genetics-based machine learning.

The compact classifier system: Motivation, analysis and first results

Xavier — Wed, 20 Jul 2005 22:13:47 +0000

by Xavier Llorà, Kumara Sastry, and David E. Goldberg (2006).
Proceedings of the Congress on Evolutionary Computation, 1, 596—603. Also as IlliGAL TR No 2005019. Link to the PDF.

Abstract
This paper presents an analysis of how maximally general and accurate rules can be evolved in a Pittsburgh-style classifier system. In order to be able to perform such an analysis we introduce a simple bare-bones Pittsburgh-style classifier systems—the compact classifier system (CCS)—based on estimation of distribution algorithms. Using a common rule encoding schemes of Pittsburgh-style classifier systems, CCS mantains a dynamic set of probability vectors that compactly describe a rule set. The compact genetic algorithm is used to evolve each of the initially perturbated probability vectors. Results show how CCS is able to evolve in a compact, simple, and elegant manner rule sets composed by maximally general and accurate rules. The initial theoretical analysis and results also show that traditional encoding schemes used by Pittsburgh-style classifiers add an extra facet of diffiiculty. Such a bias plays a central role on the overall performance and scalability of CCS and other Pittsburgh-style systems using such encoding schemes.