"A number of representation schemes have been presented for use within Learning Classifier Systems, ranging from binary encodings to neural networks. This paper presents results from an investigation into using a discrete dynamical system representation within the XCS Learning Classifier System. In particular, asynchronous random Boolean networks are used to represent the traditional condition-action production system rules. It is shown possible to use self-adaptive, open-ended evolution to design an ensemble of such discrete dynamical systems within XCS to solve a number of well-known test problems."

"A number of representation schemes have been presented for use within Learning Classifier Systems, ranging from binary encodings to Neural Networks, and more recently Dynamical Genetic Programming (DGP). This paper presents results from an investigation into using a fuzzy DGP representation within the XCSF Learning Classifier System. In particular, asynchronous Fuzzy Logic Networks are used to represent the traditional condition-action production system rules. It is shown possible to use self-adaptive, open-ended evolution to design an ensemble of such fuzzy dynamical systems within XCSF to solve several well-known continuous-valued test problems."

"A number of representation schemes have been presented for use within Learning Classifier Systems, ranging from binary encodings to neural networks. This paper presents results from an investigation into using discrete and fuzzy dynamical system representations within the XCSF Learning Classifier System. In particular, asynchronous Random Boolean Networks are used to represent the traditional condition-action production system rules in the discrete case and asynchronous Fuzzy Logic Networks in the continuous-valued case. It is shown possible to use self-adaptive, open-ended evolution to design an ensemble of such dynamical systems within XCSF to solve a number of well-known test problems."

The technology has been developed to withstand the challenges of real world — in addition to handling problems of too much data, too little data, correlated data, or noisy data, DataModeler respects the cost and timeliness issues associated with modeling development.

"We investigate the evolutionary dynamics of an idealised model for the robust self-assembly of two-dimensional structures called polyominoes. The model includes rules that encode interactions between sets of square tiles that drive the self-assembly process. The relationship between the model's rule set and its resulting self-assembled structure can be viewed as a genotype-phenotype map and incorporated into a genetic algorithm."

"The task of the 11-bit boolean multiplexer is to decode a 3-bit binary address (000, 001, 010, 011, 100, 101, 110, 111) and return the value of the corresponding data register (d0, d1, d2, d3, d4, d5, d6, d7). Thus, the boolean 11-multiplexer is a function of 11 arguments: three, a0 to a2, determine the address, and eight, d0 to d7, determine the answer. As GEP uses single-character chromosomes, T = {a, b, c, 1, 2, 3, 4, 5, 6, 7, 8} which correspond, respectively, to {a0, a1, a2, d0, d1, d2, d3, d4, d5, d6, d7}."

"The three-dimensional solar structure could, at least in principle, absorb a lot more light and generate more power than a flat panel containing the same area footprint. The hope is that all unused light which has been reflected off one panel would be captured by other panels. Panels of this type would be most ideal in circumstances with limited space."

can;t actually read this :/

Want to breed replacements for the models in Section 4

"The use of Graphics Processing Units (GPUs) in scientific computing is becoming increasingly common. GPUs are low cost parallel processors that can readily be exploited for many types of general purpose computation. Recently, the computational intelligence community has started to develop for the GPU platform. This web page is primarily dedicated to the use of GPUs as a platform for Genetic Programming. "

"The mean spectra were provided as input sequences to the Implicit Context Representation Cartesian Genetic Programming algorithm (IRCGP)[14,15]. IRCGP uses evolutionary computing methodology to learn classifiers that are capable of distinguishing between data classes. Induced classifiers take the form of programmatic expressions applied to particular offsets within the input data sequences. These expressions are composed from a set of simple mathematical functions. Both the choice and connectivity of the functions, and the choice of offsets used within the input sequences, are determined by the algorithm's evolutionary process. The input sequences were divided equally into training and test sets. To prevent over-learning, training of the classifiers was stopped once classification accuracy of the test sequences started to fall."

"The fourth issue of volume 10 of Genetic Programming and Evolvable Machines is now available online. This is the first part of the two-part Special Issue on Parallel and Distributed Evolutionary Algorithms, and it contains the following articles:..." [which I unfortunately cannot read; dammit, Springer]

"Many GECCO papers discuss lessons learned in a particular application, but few papers discuss lessons learned over an ensemble of problem areas. A scan of the tables of contents of the Proceedings from GECCO 2005 and 2006 showed no paper title stressing lessons learned although the term "pitfall" appeared occasionally in abstracts, typically applying to a particular practice. We present in this paper a set of broadly applicable "lessons learned" in the application of evolutionary computing (EC) techniques to a variety of problem areas and present advice related to encoding, running, monitoring, and managing an evolutionary computing task."

"This paper has two aspects. First, it describes the use of genetic programming to automatically synthesize a solution to the challenge problem posed at an international competition held every four years in the field of optical design. In 2002, the competition at the International Optical Design Conference attracted 42 entries from 39 well-known optical designers, commercial consultants, and patent holders from many of the field's most prominent companies, universities, and research institutions. The 39 human contestants spent an average of 34.1 hours working on their entries. Virtually all entries were considered good solutions to the challenge problem. Genetic programming automatically synthesized a design "from scratch" - that is, without starting from a pre-existing human-created design and without pre-specifying the number of lenses, the physical layout of the lenses, or the numerical or non-numerical parameters of the lenses...."

"About the Book: This is an open set of lecture notes on metaheuristics algorithms, intended for undergraduate students, practitioners, programmers, and other non-experts. It was developed as a series of lecture notes for an undergraduate course I taught at GMU. The chapters are designed to be printable separately if necessary. As it's lecture notes, the topics are short and light on examples and theory. It's best when complementing other texts. With time, I might remedy this."

See in particular Chapter 2, on Data Balancing. This is important stuff for those of us dealing with data-driven models and techniques, especially those not based on analytical closed form first-principles junk.

"Genetics Squared's test would be able to tell which category the patients fit into, potentially saving hospitals loads of money in unnecessary treatment and patients debilitating chemotherapy. The company hopes to begin marketing the test by the third quarter of this year."

Actually, this is a form of linear genetic programming (GP). A "genetic algorithm" typically refers to mapping genes directly onto a fixed series of parameters. Genetic programming refers to search over the set of arbitrary-size executable code, or structures of arbitrary complexity.

Interesting prospect for a Nudge application

"Last year, more than 16,000 companies filed 10-Ks or 10KSBs with the SEC. Assuming they come in evenly (they don't, but we'll say this for simplicity's sake), that would be more than 4,000 annual reports a quarter, or roughly 44 a day, each and every day of the year. Forget holidays, vacations, or your kids' birthdays — you've got annual reports to read!"

We were talking about the Uncanny Valley a few days ago, and I was reminded of Hod's dreaming spider robots, twitching in their sleep.

One of the reasons I've been kind of quiet lately.

Applying an EDA to linear GP. As it happens, Push is a (roughly) linear GP language.

Considering possibility of submitting some MSS

Javascript implementation of [most of] Push, suitable for inclusion in unwitting distributed computational systems.

Just musing about a fun genetic programming project

I love this. Love it. Lee and Jon, you made my day.

Human-competitive automated design competition, using genetic programming. Some interesting entries this year.

Historical market data source.

Several genetic programming folks were asking after "real world" datasets last week. Here some are.

I'll be there

Seems to have been updated to include more recent Humies.