to search for graphs that exhibit one of two extreme behaviors: maximum epidemic duration or maximal number of individuals catching the disease. The focus of the paper is on comparison of two representations for evolvable networks. The first makes local expansions of the network specified by a linear

-forward perceptrons with a single hidden layer. The population size and the number of nodes in the hidden layer are varied across a series of experiments. The results of the simulations indicate a minimum amount of complexity is required in a player’s strategy in order for cooperation to evolve. Moreover, under

The authors investigate the use of a genetic algorithm to control the evolution of artificial neural networks for the purpose of detecting edges in single-line digitized images. Specialpurpose neurons are used in an architecture which is fixed in terms of inter-layer connectivity but free in terms of number of layers, activation functions, and other attributes. Preliminary results indicate that three of the free attributes do have a best form. I.

of elements. We hypothesize that FLN only includes recursion and is the only uniquely human component of the faculty of language. We further argue that FLN may have evolved for reasons other than language, hence comparative studies might look for evidence of such computations outside of the domain

to evolve optimal parameters for an exploratory swarm. KEY WORDS: swarm behavior, distributed control, evolutionary optimization, genetic algorithms 1.

Existing visualization tools typically do not allow easy extension by new visualization techniques, and are often coupled with inflexible data input mechanisms. This paper presents EVolve, a flexible and extensible framework for visualizing program characteristics and behaviour. The framework

, in which the genotypes are treated as genetic regulatory networks. The dynamics of the regulatory network direct the growth of the agent, and lead to the construction of both the morphology and neural control of the agent. We demonstrate that such a model can be used to evolve agents to perform non

Systems which evolve music are becoming increasingly popular under the domain of interactive evolution. Steady state neuroevolution can be used for both learning an example and learning through interaction. This provides an avenue to creative composition. However, too much freedom in creation can

Recent research in automated highway systems has ranged from low-level vision-based controllers to highlevel route-guidance software. However there is currently no system for tactical-level reasoning. Such a system should address tasks such as passing cars, making exits on time, and merging into a traffic stream. Our approach to this intermediate-level planning combines a distributed reasoning system (PolySAPIENT) with a novel evolutionary optimization strategy (PBIL). PBIL automatically tunes PolySAPIENT module parameters in simulation by evaluating candidate modules on various traffic scenarios. Since the control interface to the simulated vehicles is identical to that on the Carnegie Mellon Navlab vehicles, modules developed using this process can be directly ported to existing hardware. This method is currently being applied to the automated highway system domain; it also generalizes to many complex robotics tasks where multiple interacting modules must simultaneously be configured without individual module feedback. 1.

Software systems can be developed through evolution (gradual change) or revolution (reimplementation from scratch). Both approaches have advantages and disadvantages. An evolutionary approach keeps the system working throughout, allowing early problem detection, but tends to retain ingrained design flaws and can result in complex, ad hoc systems. A revolutionary approach is required to change the basic architecture of a system, but many more resources must be invested before the system can be evaluated. In this paper, we describe how we used a little application-specific language to combine these approaches ’ advantages. The context of our work is CTAS [2], the nextgeneration air traffic control automation system developed originally by NASA. The overall goal was to redesign and reimplement one of the CTAS processes in Java, while retaining its ability to communicate with unmodified processes—a project complicated by CTAS’s ad hoc message formats. To address this, we designed a language that combines C code copied from CTAS source, to express the message formats, with new Java code for message actions. A compiler then automatically generates code for marshalling and unmarshalling. The result is a system with both evolutionary and revolutionary properties, exemplified by the use of both old CTAS code and new Java code in the message language. This paper discusses the language and compiler and evaluates some of the engineering tradeoffs inherent in their design.