Alfonso Ortega de la Puente

Attribute grammar evolution

This paper describes Attribute Grammar Evolution (AGE), a new Automatic Evolutionary Programming algorithm that extends standard Grammar Evolution (GE) by replacing context-free grammars by attribute grammars. GE only takes into account syntactic restrictions to generate valid individuals. AGE adds semantics to ensure that both semantically and syntactically valid individuals are generated. Attribute grammars make it possible to semantically describe the solution. The paper shows empirically that AGE is as good as GE for a classical problem, and proves that including semantics in the grammar can improve GE performance. An important conclusion is that adding too much semantics can make the search difficult.

Automatic composition of music by means of grammatical evolution

This work describes how grammatical evolution may be applied to the domain of automatic composition. Our goal is to test this technique as an alternate tool for automatic composition. The AP440 auxiliary processor will be used to play music, thus we shall use a grammar that generates AP440 melodies. Grammar evolution will use fitness functions defined from several well-known single melodies to automatically generate AP440 compositions that are expected to sound like those composed by human musicians.

Parallel Simulation of NEPs on Clusters

This paper compares two different approaches, followed by our research group, to efficiently run NEPs on parallel platforms, as general and transparent as possible. The vague results of jNEP (our multithreaded Java simulator for multicore desktop computers) suggests the use of massively parallel platforms (clusters of computers). The good results obtained show the scalability and viability of this last approach.

Distributed Simulation of NEPs Based On-Demand Cloud Elastic Computation

Networks of Evolutionary Processors (NEP) are a bio-inspired computational model able to solve NP complete problems in an efficient manner. Up to now, the only way to analyze and execute these devices is through hardware and software simulators able to encapsulate the inherent parallelism and the efficiency in their computations. Nowadays, simulators for these models only cover many software applications developed under sequential/parallel architectures over multicore desktop computers or clusters of computers. Most of them, are not able to handle the size of non trivial problems within a massively parallel environment. We consider that cloud computation offers an interesting and promising option to overcome the drawbacks of these solutions. In this paper, we propose a novel parallel distributed architecture to simulate NEPs using on-demand cloud elastic computation. A flexible and extensible simulator is developed in order to demonstrate the suitability and scalability of our architecture with several variants of NEP.

A Christiansen Grammar for Universal Splicing Systems

The main goal of this work is to formally describe splicing systems. This is a necessary step to subsequently apply Christiansen Grammar Evolution (an evolutionary tool developed by the authors) for automatic designing of splicing systems. Their large number of variants suggests us a decisions: to select a family as simple as possible of splicing systems equivalent to Turing machines. This property ensures that the kind of systems our grammar can generate is able to solve any arbitrary problem. Some components of these universal splicing systems depend on other components. So, a formal representation able to handle context dependent constructions is needed. Our work uses Christiansen grammars to describe splicing systems.

Towards the automatic programming of NEPs

This paper shows the platform with which we implement a general methodology to automatically design NEPs to solve specific problems. We use CGE/AGE (a new genetic programming algorithm) and jNEP (a Java NEP simulator), two applications we have previously developed. This work is just a proof of viability. We are interested on linking all the modules and generating the initial population. Building this platform is relevant, because our methodology includes several non trivial steps, such as designing a grammar, and implementing and using a simulator. For this first proof we have choosen a well known problem that other authors have solved by means of NEPs.

Towards Quantitative Networks of Polarized Evolutionary Processors: A Bio-Inspired Computational Model with Numerical Evaluations

Networks of Polarized Evolutionary Processors is a highly parallel distributed computing model inspired and abstracted from the biological evolution. This model is computationally complete and able to efficiently solve NP complete problems. Although this model is inspired from biology, basically it has been investigated from the points of view of mathematical and computer science goals with a qualitative perspective. It is true that Networks of Polarized Evolutionary Processors incorporate a numerical evaluation over the data that it processes, but this is not used from a quantitative viewpoint. In this paper we propose to enhance Networks of Polarized Evolutionary Processors of a quantitative perspective through a novel number of formal components. In particular, these components are able to evaluate quantitative conditions inherent to biological phenomena preserving the same computational power of Networks of Polarized Evolutionary Processors. Moreover, as a proof of concept, we model and simulate a simple but expressive example: a discrete abstraction of the sodium-potassium pump that includes the components proposed. Finally, we suggest that this integration enhances Networks of Polarized Evolutionary Processors model to (a) be more expressive for the algorithm design and (b) use less resources (nodes, rules, strings and computation time). This resource reduction could become a clear advantage when we will deploy hardware/software solutions of these bio-inspired computational models on top of massively distributed computational platforms.

Model Regularization in Coevolutionary Architectures Evolving Straight Line Code

Frequently, when an evolutionary algorithm is applied to a population of symbolic expressions, the shapes of these symbolic expressions are very different at the first generations whereas they become more similar during the evolving process. In fact, when the evolutionary algorithm finishes most of the best symbolic expressions only differ in some of its coefficients. In this paper we present several coevolutionary strategies of a genetic program that evolves symbolic expressions represented by straight line programs and an evolution strategy that searches for good coefficients. The presented methods have been applied to solve instances of symbolic regression problem, corrupted by additive noise. A main contribution of the work is the introduction of a fitness function with a penalty term, besides the well known fitness function based on the empirical error over the sample set. The results show that in the presence of noise, the coevolutionary architecture with penalized fitness function outperforms the strategies where only the empirical error is considered in order to evaluate the symbolic expressions of the population.