<?xml version="1.0" encoding="UTF-8"?><xml><records><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Maciej Komosinski</style></author><author><style face="normal" font="default" size="100%">Konrad Miazga</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Comparison of the tournament-based convection selection with the island model in evolutionary algorithms</style></title><secondary-title><style face="normal" font="default" size="100%">Journal of Computational Science</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2018</style></year></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://www.framsticks.com/files/common/ConvectionSelectionVsIslandModel.pdf</style></url></web-urls></urls><volume><style face="normal" font="default" size="100%">32</style></volume><pages><style face="normal" font="default" size="100%">106–114</style></pages><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">Convection selection is an approach to multipopulational evolutionary algorithms where solutions are assigned to subpopulations based on their fitness values. Although it is known that convection selection can allow the algorithm to find better solutions than it would be possible with a standard single population, the convection approach was not yet compared to other, commonly used architectures of multipopulational evolutionary algorithms, such as the island model. In this paper we describe results of experiments which facilitate such a comparison, including extensive multi-parameter analyses. We show that approaches based on convection selection can obtain better results than the island model, especially for difficult optimization problems such as those existing in the area of evolutionary design. We also introduce and test a generalization of the convection selection which allows for adjustable overlapping of fitness ranges of subpopulations; the amount of overlapping influences the exploration vs. exploitation balance.</style></abstract></record><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>47</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Andrzej Gajda</style></author><author><style face="normal" font="default" size="100%">Adam Kups</style></author><author><style face="normal" font="default" size="100%">Mariusz Urbański</style></author></authors><secondary-authors><author><style face="normal" font="default" size="100%">M. Ganzha</style></author><author><style face="normal" font="default" size="100%">L. Maciaszek</style></author><author><style face="normal" font="default" size="100%">M. Paprzycki</style></author></secondary-authors></contributors><titles><title><style face="normal" font="default" size="100%">A connectionist approach to abductive problems: employing a learning algorithm</style></title><secondary-title><style face="normal" font="default" size="100%">Proceedings of the 2016 Federated Conference on Computer Science and Information Systems (FedCSIS)</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2016</style></year></dates><publisher><style face="normal" font="default" size="100%">ACSIS</style></publisher><pages><style face="normal" font="default" size="100%">353–362</style></pages><abstract><style face="normal" font="default" size="100%">This paper presents preliminary results of an application of artificial neural networks and Backpropagation learning algorithm to solve logical abductive problems. To represent logic programs in the form of artificial neural networks CIL2P approach proposed by Garcez et al. is employed. Our abductive procedure makes use of translation of a logic program representing a knowledge base into a neural network, training of the neural network with an example representing an abductive goal and translation of the trained network back to the form of a logic program. An abductive hypothesis is represented as the symmetric difference between the initial logic program and the one obtained after training of the network. The first part of the paper introduces formal description of the tools used to model the abductive process, while the second part illustrates our contribution with results of a few computational experiments and discusses the ways of possible improvements of the proposed procedure.</style></abstract></record><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Maciej Komosinski</style></author><author><style face="normal" font="default" size="100%">Adam Rotaru-Varga</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Comparison of different genotype encodings for simulated 3D agents</style></title><secondary-title><style face="normal" font="default" size="100%">Artificial Life Journal</style></secondary-title></titles><keywords><keyword><style  face="normal" font="default" size="100%">Agents</style></keyword><keyword><style  face="normal" font="default" size="100%">AL</style></keyword><keyword><style  face="normal" font="default" size="100%">Body and Brain evol.</style></keyword><keyword><style  face="normal" font="default" size="100%">Genetics</style></keyword><keyword><style  face="normal" font="default" size="100%">Simulation</style></keyword><keyword><style  face="normal" font="default" size="100%">Theory</style></keyword></keywords><dates><year><style  face="normal" font="default" size="100%">2001</style></year><pub-dates><date><style  face="normal" font="default" size="100%">Fall</style></date></pub-dates></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://www.framsticks.com/files/common/ComparisonGeneticEncodings3DAgents.pdf</style></url></web-urls></urls><number><style face="normal" font="default" size="100%">4</style></number><publisher><style face="normal" font="default" size="100%">MIT Press</style></publisher><pub-location><style face="normal" font="default" size="100%">Cambridge, MA</style></pub-location><volume><style face="normal" font="default" size="100%">7</style></volume><pages><style face="normal" font="default" size="100%">395–418</style></pages><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">This paper analyzes the effect of different genetic encodings used for evolving 3D agents with physical morphologies. The complex phenotypes used in such systems often require nontrivial encodings. Different encodings used in Framsticks - a system for evolving 3D agents - are presented. These include a low-level direct mapping and two higher-level encodings: a recurrent and a developmental one. Quantitative results are presented from three simple optimization tasks (active height, passive height, and locomotion speed). The low-level encoding produced solutions of lower fitness than the two higher-level encodings under similar conditions. Results from recurrent and developmental encodings had similar fitness values but displayed qualitative differences. Desirable advantages and some drawbacks of more complex encodings are established.</style></abstract></record></records></xml>