<?xml version="1.0" encoding="UTF-8"?><xml><records><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%">Kaszuba, Piotr</style></author><author><style face="normal" font="default" size="100%">Komosinski, Maciej</style></author><author><style face="normal" font="default" size="100%">Mensfelt, Agnieszka</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Automated development of latent representations for optimization of sequences using autoencoders</style></title><secondary-title><style face="normal" font="default" size="100%">2021 IEEE Congress on Evolutionary Computation (CEC)</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2021</style></year></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://www.framsticks.com/files/common/LatentRepresentationsForSequencesOptimization.pdf</style></url></web-urls></urls><publisher><style face="normal" font="default" size="100%">IEEE</style></publisher><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">In this paper, we propose an automated method for the development of new representations of sequences. For this purpose, we introduce a two-way mapping from variable length sequence representations to a latent representation modelled as the bottleneck of an LSTM (long short-term memory) autoencoder. Desirable properties of such mappings include smooth fitness landscapes for optimization problems and better evolvability. This work explores the capabilities of such latent encodings in the context of optimization of 3D structures. Various improvements are adopted that include manipulating the autoencoder architecture and its training procedure. The results of evolutionary algorithms that use different variants of automatically developed encodings are compared.</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%">Komosinski, Maciej</style></author><author><style face="normal" font="default" size="100%">Miazga, Konrad</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Diversity control in evolution of movement</style></title><secondary-title><style face="normal" font="default" size="100%">Artificial Life Conference Proceedings</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2021</style></year></dates><publisher><style face="normal" font="default" size="100%">MIT Press</style></publisher><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">In this work we investigate how various techniques of diversity control employed during evolution of 3D agents influence the velocity they achieve, and how these techniques influence the diversity of behaviors across multiple independent evolutionary runs. Three evolutionary settings are compared: a standard generational evolutionary process where fitness is velocity, a niching technique, and pure novelty search. Two genetic encodings (lower and higher level) and two environments (land and water) are used in experiments. To diversify behaviors, seven properties of movement introduced earlier are calculated for each individual during evolution. Best individuals obtained from evolution in each setting are compared both in terms of their fitness and the similarity of their movement patterns.</style></abstract></record><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>27</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Komosinski, Maciej</style></author><author><style face="normal" font="default" size="100%">Kups, Adam</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Type A and Type B Effects, Time-Order Error and Weber's Law in Human Timing – Simulations and Synthesis</style></title></titles><dates><year><style  face="normal" font="default" size="100%">2021</style></year></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://www.framsticks.com/files/common/HumanTimingSimulation-TypeA-TypeB-TOE-WebersLaw.pdf</style></url></web-urls></urls><publisher><style face="normal" font="default" size="100%">Poznan University of Technology, Institute of Computing Science</style></publisher><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">This article presents a computational approach to the theoretical integration of the psychophysical phenomena in human timing. While there are many useful models of human timing, analyses are scarce on how these models explain the relationships between several phenomena at the same time. The presented research is an attempt to primarily explain and integrate the time-order error with the Type A and Type B phenomena. The final result of this work also encompasses Weber's law property and relates it to the aforementioned order-related effects. The theoretical framework used is the Clock-Counter Timing Network (CCTN), an artificial neural network timing model which has been constructed to explain the process of comparing durations of stimuli. Extensive simulations performed with the use of this model revealed that the considered psychophysical properties may be strongly interrelated and dependent on a simple perceptual mechanism. The obtained results allow to formulate specific experimentally testable predictions.</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%">Komosinski, Maciej</style></author><author><style face="normal" font="default" size="100%">Mensfelt, Agnieszka</style></author></authors><secondary-authors><author><style face="normal" font="default" size="100%">Kaufmann, Paul</style></author><author><style face="normal" font="default" size="100%">Castillo, Pedro A.</style></author></secondary-authors></contributors><titles><title><style face="normal" font="default" size="100%">A Flexible Dissimilarity Measure for Active and Passive 3D Structures and Its Application in the Fitness–Distance Analysis</style></title><secondary-title><style face="normal" font="default" size="100%">Applications of Evolutionary Computation</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2019</style></year></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://www.framsticks.com/files/common/DissimilarityMeasure3DStructuresFitnessDistance.pdf</style></url></web-urls></urls><publisher><style face="normal" font="default" size="100%">Springer</style></publisher><isbn><style face="normal" font="default" size="100%">978-3-030-16692-2</style></isbn><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">Evolutionary design of 3D structures – either static structures, or equipped with some sort of a control system – is one of the hardest optimization tasks. One of the reasons are rugged fitness landscapes resulting from complex and non-obvious genetic representations of such structures and their genetic operators. This paper investigates global convexity of fitness landscapes in optimization tasks of maximizing velocity and height of both active and passive structures. For this purpose, a new dissimilarity measure for 3D active and passive structures represented as undirected graphs is introduced. The proposed measure is general and flexible – any vertex properties can be easily incorporated as dissimilarity components. The new measure was compared against the previously introduced measure in terms of triangle inequality satisfiability, changes in raw measure values and the computational cost. The comparison revealed improvements for triangle inequality and raw values at the expense of increased computational complexity. The investigation of global convexity of the fitness landscape, involving the fitness–distance correlation analysis, revealed negative correlation between the dissimilarity of the structures and their fitness for most of the investigated cases.</style></abstract></record></records></xml>