ABSTRACT
As artificial life systems grow in number and sophistication, it is becoming increasingly important that the field agree on principled metrics for evaluating them. This report describes a series of experiments validating the evolutionary activity statistics developed by Bedau and his colleagues [2, 3, 4]. The work described herein was motivated by a feeling that the 'null hypothesis'---that is, that the evolutionary activity statistics fail to exclude intuitively unlifelike systems from Class 3 dynamics [3]---had not been sufficiently disproved in the existing literature. We conducted a series of experiments applying the statistics to such systems, attempting to 'break' the scheme by measuring Class 3 dynamics in an intuitively unlifelike system. The evolutionary activity measurement scheme has so far proved robust to our attempts to break it, but we believe that this work is still valuable in advancing the validity of the scheme, and that this does not mean the scheme is without shortcomings.
- J. C. Avise. Phylogeography: The History and Formation of Species. Harvard University Press, 2000.]]Google Scholar
- M. A. Bedau and N. H. Packard. Measurement of evolutionary activity, teleology, and life. In C. Langton, C. Taylor, D. Farmer, and S. Rasmussen, editors, Artificial Life II, Santa Fe Institute Studies in the Sciences of Complexity, Volume X, pages 431--461, 1992.]]Google Scholar
- M. A. Bedau, E. Snyder, C. T. Brown, and N. H. Packard. A comparison of evolutionary activity in artificial evolving systems and in the biosphere. In P. Husbands and I. Harvey, editors, Proceedings of the Fourth European Conference on Artificial Life, pages 125--134. MIT Press, 1997.]]Google Scholar
- M. A. Bedau, E. Snyder, and N. H. Packard. A classification of long-term evolutionary dynamics. In C. Adami, R. Belew, H. Kitano, and C. Taylor, editors, Artificial Life VI, pages 228--237. MIT Press, 1998.]] Google ScholarDigital Library
- A. Channon. Passing the ALife test: Activity statistics classify evolution in Geb as unbounded. In Proceedings of the European Conference on Artificial Life, 2001.]] Google ScholarDigital Library
- A. Channon. Improving and Still Passing the ALife Test: Component-Normalised Activity Statistics Classify Evolution in Geb as Unbounded. In Standish, Abbass, and Bedau, editors, Artificial Life VIII, pages 173--181. MIT Press, 2002.]] Google ScholarDigital Library
- S. J. Gould. The Structure of Evolutionary Theory. Belknap Press, 2002.]]Google ScholarCross Ref
Index Terms
- Validation of evolutionary activity metrics for long-term evolutionary dynamics
Recommendations
Diversity Guided Evolutionary Programming: A novel approach for continuous optimization
Avoiding premature convergence to local optima and rapid convergence towards global optima has been the major concern with evolutionary systems research. In order to avoid premature convergence, sufficient amount of genetic diversity within the evolving ...
An evolutionary benefit from misperception in foraging behaviour
ACAL'07: Proceedings of the 3rd Australian conference on Progress in artificial lifeMisperception is a common cause of error for individuals and organisations. Conventional wisdom suggests that its effects are detrimental to the misperceiver or its society as a whole. However, in some circumstances misperception can provide a benefit ...
Diversity metrics for direct-coded variable-length chromosome shortest path problem evolutionary algorithms
AbstractOne of the major concerns in Evolutionary Algorithms is the premature convergence caused by what is known as the Exploration and Exploitation Balance problem. To maintain this balance, population diversity should be maintained during the ...
Comments