Artificial Evolution

Artificial Life (A-Life) and Evolutionary Algorithms (EA) provide a variety of new techniques for making and studying music. EA have been used in different musical applications, ranging from new systems for composition and performance, to models for studying musical evolution in artificial societies. This paper starts with a brief introduction to three main fields of application of EA in Music, namely sound design, creativity and computational musicology. Then it presents our work in the field of computational musicology. Computational musicology is broadly defined as the study of Music with computational modelling and simulation. We are interested in developing A-Life-based models to study the evolution of musical cognition in an artificial society of agents. In this paper we present the main components of a model that we are developing to study the evolution of musical ontogenies, focusing on the evolution of rhythms and emotional systems. The paper concludes by suggesting that A-Life and EA provide a powerful paradigm for computational musicology.

This paper reports on investigations on the possible advantage of the coupling between genomes and physics of cells in artificial evolution. The idea is simple: evolution can rely on physical processes during development allowing to produce shapes without need to specify how exactly this shaping has to be done. Evolving a minimal energy surface such as soap bubbles would need only the specification of the boundary values and a homogenous interaction pattern between the cells. This paper shows that it is possible to link a genetic regulatory network to physics during development, that a reduction of parameters is indeed possible and that the understanding of what is going on in such a system is relatively easy to gain.

Increasingly often artificial evolutionary techniques are coupled with mechanism abstracted from developmental biology. For instance, artificial cells endowed with genetic regulatory networks were used to evolve and develop simulated creatures. With the evolution of a simple vermicular structure it is shown that asymmetric cell division is useful for the positioning of cells and that this mechanism can be integrated with other developmental mechanisms such as genetic regulation and cell adhesion to get moving artificial creatures. Surprisingly, the movementv were controlled by the genetic regulatory network alone without the need to evolve a neural structure.

In this paper we describe our ongoing work on the control of a tendon driven robotic hand by an adaptive learning mechanism evolved using a simulator developed over the last years. The proposed neural network allows the robotic hand to explore its own movement possibilities to interact with objects of different shape, size and material and learn how to grasp them. As the evolved neural controller is highly adaptive, it will allow us in the future to systematically investigate the interplay between morphology and behavior using the same, but adaptive neural controller.

In this paper we propose an ant colony optimization variant where several independent colonies try to simultaneously solve the same problem. The approach includes a migration mechanism that ensures the exchange of information between colonies and a mutation operator that aims to adjust the parameter settings during the optimization. The proposed method was applied to several benchmark instances of the node placement problem. The results obtained shown that the multi-colony approach is more effective than the single-colony. A detailed analysis of the algorithm behavior also reveals that it is able to delay the premature convergence.

The publication of the human genome has elicited commentary to the effect that since fewer genes were identified than anticipated, it follows that genes are less important to human biology than anticipated. The flaws in this syllogism are explained in the context of a treatise on how genomes operate and evolve and how genes function to produce embryos and brains. Most of our most cherished human traits are the result of the emergence of new properties from pre-existing genetically scripted ideas, offering countless opportunities to celebrate the evolutionary process.

We study the impact of backbones in optimization and approximation problems. We show that some optimization problems like graph coloring resemble decision problems, with problem hardness positively correlated with backbone size. For other optimization problems like blocks world planning and traveling salesperson problems, problem hardness is weakly and negatively correlated with backbone size, while the cost of finding optimal and approximate solutions is positively correlated with backbone size. A third class of optimization problems like number partitioning have regions of both types of behavior. We find that to observe the impact of backbone size on problem hardness, it is necessary to eliminate some symmetries, perform trivial reductions and factor out the effective problem size.

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Hidden Markov Models (HMMs) have been successfully used in tasks involving prediction and recognition of patterns in sequence data, with applications in areas such as speech recognition and bioinformatics. While variations of traditional HMMs proved to be practical in applications where it is feasible to obtain the numerical probabilities required for the specification of the parameters of the model and the probabilities available are descriptive of the underlying uncertainty, the capabilities of HMMs remain unexplored in applications where this convenience is not available. Motivated by such applications, we present a HMM that uses qualitative probabilities instead of quantitative ones. More specifically, the HMM presented here captures the order of magnitude of the probabilities involved instead of numerical probability values. We analyze the resulting model by using it to perform classification tasks on gene expression data.

As early as the 1950s and early 1960s, pioneers such as Lejaren Hiller, Gottfried Michael Koenig, Iannis Xenakis, and Pietro Grossi, among a few others, started to gain access to computers to make music. It soon became clear that to render music with a socalled "human feel," computers needed to process information about performance (e.g., deviations in tempo and loudness), in addition to the symbols that are normally found in a traditional musical score (e.g., pitch and rhythm). This was especially relevant for those interested in using the computer to play back scores. Indeed, the first ever attempt at creating a computer-music programming language, by Max Mathews at Bell Telephone Laboratories in 1957, was motivated by his wish to "write a program to perform music on the computer" (Park 2009 p. 10). It appears that this development began after Mathews and John Pierce went to a piano concert together. During the intermission, Pierce suggested that perhaps a computer could perform as well as the pianist. Mathews took up the challenge, which resulted in Music I, the ancestor of music programming languages such as Csound . Research into computational models of expressive performance of music (Widmer and Goebl 2004) is still an active area of study-particularly, research into devising increasingly more sophisticated automated and semi-automated computer systems for expressive music performance, hereinafter referred to as CSEMP. A CSEMP is able to generate expressive performances of music. For example, software for music typesetting is often used to write a piece of music, but most packages play back the music in a

Many multi-agent systems seek to reconcile two apparently inconsistent constraints. The system's overall objective is defined at a global level. However, the agents have only local information available to them in selecting their actions. Such systems are presently more art than science. They often exhibit regularities (such as exponential convergence) that we do not understand, and we do not know how to improve their functioning in a disciplined manner. In this paper, we develop a simple statistical model for such systems that can enhance both our intuitions about their functioning and our ability to engineer them, and apply it to three systems that have been constructed.

The recently released documentary titled "The Hunt for the Oldest DNA" was the inspiration for the writing of this paper. It is because Professor Eske Willerslev and I, David R. Wood, are both peers in two mirror fields of evolutionary science achieving similar breakthrough results using similar techniques to unconceal ancient DNA-natural and artificial. This paper goes through the documentary section by section to describe how Willerslev's and my journey of exploration to make breakthrough scientific discoveries follow literally the same pattern. The reason is that hunting for ancient natural DNA and artificial DNA which are variations of each other requires the same adaptations to align with the value stream demanded to make our respective scientific discoveries. And the paper concludes with a call for the integration of these two distinct research efforts in two peer fields of evolutionary science. To integrate the crossinterdisciplinary information, perspectives, and techniques to enhance both research efforts. This will predictably enable accelerated scientific discoveries in both evolutionary fields in the future.

This paper is concerned with an aspect of the design of metaheuristic algorithms, such as evolutionary algorithms, tabu search and ant colony optimization. The topic that is considered is how problems can be represented when they are given to a metaheuristic algorithm. A particular difficulty is presented, viz. the''bottleneck'', where the problem is artificially converted into a new representation in order to fit the standard input to the metaheuristic. Such bottlenecks cause problems in interpreting or trusting the solution ...

Arti cial evolution can automatically derive the con guration of a recon gurable hardware system such t h a t i t performs a given task. Individuals of the evolving population are evaluated when instantiated as real circuits, so if constraints inherent t o h uman design (but not to evolution) are dropped, then the natural physical dynamics of the hardware can be exploited in new ways. The notion of an arti cially evolving `species' (SAGA) allows the open-ended incremental evolution of complex circuits. Theoretical arguments are given, as well as the real-world example of an evolved hardware robot controller.

Arti cial evolution as a design methodology for hardware frees many of the simplifying constraints normally imposed to make d esign by h umans tractable. However, this freedom comes at some cost, and a whole fresh set of issues must be considered. Standard genetic algorithms are not generally appropriate for hardware evolution when the number of components need not be predetermined. The use of simulations is problematic, and robustness in the presence of noise or hardware faults is important. We p r e s e n t theoretical arguments, and illustrate with a physical piece of hardware evolved in the real-world (`intrinsically evolved' hardware). A simple asynchronous digital circuit controls a real robot, using a minimal sensorimotor control system of 32 bits of RAM and a few ip-ops to co-ordinate sonar pulses and motor pulses with no further processing. This circuit is tolerant to single-stuck-at faults in the RAM. The methodology is applicable to many t ypes of hardware, including Field-Programmable Gate Arrays (FPGA's).

Arti cial evolution is discussed in the context of a successful experiment e v olving a hardware con guration for a silicon chip (a Field Programmable Gate Array) the real chip was used to evaluate individual con gurations on a tone-recognition task. The evolutionary pathway i s analysed it is shown that the population is genetically highly converged and travels far through genotype space. Species Adaptation Genetic Algorithms (SAGA) are appropriate for this ty p e o f e v olution, and it is shown how an appropriate mutation rate was chosen. The role of junk on the genotype is discussed, and it is suggested that neutral networks (paths through genotype space via mutations which l e a ve t n e s s u nchanged) may be crucial to the e ectiveness of evolution.

Intrinsic' Hardware Evolution is the use of arti cial evolution | such as a Genetic Algorithm | to design an electronic circuit automatically, where each tness evaluation is the measurement of a circuit's performance when physically instantiated in a real recon gurable VLSI chip. This paper makes a detailed case-study of the rst such application of evolution directly to the con guration of a Field Programmable Gate Array (FPGA). Evolution is allowed to explore beyond the scope of conventional design methods, resulting in a highly e cient circuit with a richer structure and dynamics and a greater respect for the natural properties of the implementation medium than is usual. The application is a simple, but not toy, problem: a tone-discrimination task. Practical details are considered throughout.

The natural species homo sapiens are not a cultural species. Homo sapiens instead artificially segregates itself into many artificial species (i.e., cultures) for competitive advantage in natural intraspecies competition-warfare, economics, etc. These artificial species are defined and categorized based on the distinct combination of artificial genomes, artificial structural adaptations, and artificial behavioral adaptations. Homo sapiens organize collectively as it typically affords the majority of the natural organisms enhanced access to scarce evolutionary resources in the natural struggle for existence. And the majority of homo sapiens artificial adaptations align with the two-core natural evolutionary value streams-survival and reproduction. This is the case because all artificial adaptations produced by homo sapiens are logically an extension of natural evolution itself. This is so because the mechanism of natural selection draws no practical distinction between natural and artificial adaptations at the point of selection. The failed science of cultural evolution does not frame the artificial adaptations produced by homo sapiens in this practical way. This is why cultural evolutionary science has been unable to integrate successfully with the practical patterns of natural species, to include homo sapiens, in the fields that surround it academically. This makes the field of cultural evolution fatally flawed necessitating a replacement of their theoretical framework. My discovery of The Unified Framework of Evolution-Natural & Artificial is the Kuhnsian paradigm shift that will replace their failed framework. A unified framework that has successfully shown to have both explanatory and predictive value for the patterns of human competition in warfare, politics, business, etc. And this has been demonstrated by great theorists in their field such as Sun Tzu and Machiavelli. The field of cultural evolution must now either self-disrupt or be disrupted by this paradigm shift-or risk literal intellectual extinction of all their works.

The ideas proposed in this work are aimed to describe a novel approach based on artificial life (alife) environments for on-line adaptive optimisation of dynamical systems. The basic features of the proposed approach are: no intensive modelling (continuous learning directly from measurements) and capability to follow the system evolution (adaptation to environmental changes). The essence could be synthesized in this way: "not control rules but autonomous structures able to dynamically adapt and generate optimisedcontrol rules". We tested the proposed methodology on two applications, the Chua's circuit and a combustion process in industrial incinerators which is being carried out. Experimentation concerned the on-line optimisation and adaptation of the process in different regimes without knowing the system equations and considering one parameter affected by unknown changes. Then we let the 'alife' environment try to adapt to the new condition. Preliminary results show the system is able to dynamically adapt to slow environmental changes by recovering and tracking the optimal conditions.

running a genetic algorithm entails setting a number of parameter values. Finding settings that work well on one problem is not a trivial task and a genetic algorithm performance can be severely impacted. Moreover we know that in natural environments population sizes, reproduction and competition rates, change and tend to stabilise around appropriate values according to some environmental factors. This paper deals with a new technique for setting the genetic parameters during the course of a run by adapting the population size and the operators rates on the basis of the environmental constrain of maximum population size. In addition genetic operators are seen as alternative reproduction strategies and fighting among individuals is introduced. The algorithm is a particular instance of a chaotic map similar to the logistic map

In this paper, we detail a genetic encoding for NARS. We discuss how the method can be used to evolve instinctual knowledge and goals, and control-related parameters such as the NARS personality parameters, resulting in various NARS species. We then consider how the encoding might function in the context of a naturalistic ecosystem, wherein multiple NARS agents are embodied in simple robots, placed together in a contained environment, and evolved for animalistic tasks like eating food and reproducing. We provide some preliminary experimental observations in this direction.

The Unified Framework for the Science of Evolution. This completes the quest that began with Aristotle, progressed by Darwin, and not finished by David R. Wood. The cultural evolutionary scientific community never thought to look to existing commerical and military best practices to find the conceptual patterns they needed to complete the second process of evolution - the process of artificial evolution by means of artificial selection. Now, a former infantry officer and ITIL Master has conclusively completed this quest for the ages. It required the paradigm shift Thomas Kuhn envisioned of a resurrected ancient Greco-Roman worldview possessed by a Philosopher-Warrior to make this historic and transformational scientific discovery.

The following is a direct response to the specific work cited below and all their other academic texts that address the concept of cumulative cultural evolution. "Understanding Cumulative Cultural Evolution" published in 2016 on Academia.Edu.

In the era of commonly available problem-solving tools for, it is especially important to choose the best available method. We use Local Optima Network analysis and machine learning to select appropriate algorithm on the instance-to-instance basis. The preliminary results show that such method can be successfully applied for sufficiently distinct instances and algorithms.

Saving energy is a critical issue for mini and micro-UAVs. We used tools rooted in the 'animat' approach to generate energy saving behaviors for a glider robot. The connection weights of feed-forward neural networks were optimized by evolutionary algorithms to exhibit "soaring" behaviors, i.e. behaviors that capitalize on aerological conditions to extract energy from the environment, focusing on thermal and slope wind exploitation. Thermal soaring with spiral trajectories and slope soaring with eight-shaped trajectories were thus exhibited. The optimization criterion used for thermal soaring was the average altitude gain. For slope soaring, an additional criterion forced the glider to remain in a limited area. These criteria were high-level specifications of the desired behaviors and did not include any direct description of the strategy needed to get them.

We study the difficulty of solving different bi-objective formulations of the permutation flowshop scheduling problem by adopting a fitness landscape analysis perspective. Our main goal is to shed the light on how different problem features can impact the performance of Pareto local search algorithms. Specifically, we conduct an empirical analysis addressing the challenging question of quantifying the individual effect and the joint impact of different problem features on the success rate of the considered approaches. Our findings support that multi-objective fitness landscapes enable to devise sound general-purpose features for assessing the expected difficulty in solving permutation flowshop scheduling problems, hence pushing a step towards a better understanding of the challenges that multi-objective randomized search heuristics have to face.

This paper provides an overview of the SWARM-BOTS project, a robotic project sponsored by the Future and Emerging Technologies program of the European Commission. The paper illustrates the goals of the project, the robot prototype and the 3D simulator we built. It also reports on the results of experimental work in which distributed adaptive controllers are used to control a group of real, or simulated, robots so that they perform a variety of tasks which require cooperation and coordination. Marco Dorigo et al.

We study the evolution of social behaviors within a behavioral framework. To this end, we de ne a \minimal social situation" that is experimented with both humans and simulations based on reinforcement learning algorithms. We analyse the dynamics of behaviors in this situation by way of operant conditioning. We show that the best reinforcement algorithm, based on Staddon-Zhang's equations, has a performance and a variety of behaviors that comes close to that of humans, and clearly outperforms the well-known Q-learning. Though we use here a rather simple, yet rich, situation, we argue that operant conditioning deserves much study in the realm of arti cial life, being too often misunderstood, and confused with classical conditioning.

An interpretation of the evolution of complexity in the Iterated Prisoner's Dilemma (IPD) is developed, based on Ashby's \law of requisite variety". It is demonstrated that the in uence of noise on the evolutionary dynamics of this system is critically dependent on the locus of this noise. It is also argued that noise in such an evolving system is not merely, (or necessarily) a source of variation that must be adapted to, but, in certain circumstances, can facilitate the evolutionary exploration of increased areas of genotype space. This convergence between arti cial life/game theory and cybernetics holds implications for how we understand the general relationship between environmental complexity and agent complexity, and for our understanding of the role of noise in evolving systems.

We consider the issue of how a flexible musical space can be manipulated by users of an active music system. The musical space is navigated within by selecting transitions between different sections of the space. We take inspiration from pheromone trails in ant colonies to propose and investigate an approach that allows an artificial agent to navigate such musical spaces in accordance with the preferences of the user, and a set of boundaries specified by the designer of the musical space.

We propose an evolving ecosystem approach to evolving complex agent behaviour based on the principle of natural selection. The agents start with very limited functional design and morphology and neural controllers are concurrently evolved as functional wholes. The agents are 'grounded' in an increasingly complex environment by a complex model metabolism and interaction dynamics. Furthermore, we introduce a novel criterion for evaluating differential reproductive success aimed at maximising evolutionary freedom. We also present first experimental results suggesting that this approach may be conducive to widening the scope of artificial evolution for the generation of agents exhibiting non-trivial behaviours in a complex ecosystem.

A new side-contacted field effect diode (S-FED) structure has been introduced as a modified S-FED, which is composed of a diode and planar double gate MOSFET. In this paper, drain current of modified and conventional S-FEDs were investigated in on-state and off-state. For the conventional S-FED, the potential barrier height between the source and the channel is observed to become larger and the flow of injected electrons is reduced. Thus, the drain current decreases in on-state. While in offstate, the potential barrier height and width become smaller in conventional S-FED and so the drain current is greater than that of modified structure. Mixed mode simulations were used to determine the performance of the proposed logic gates. We compared the operation of modified S-FED with that of conventional S-FED. Simulated power delay product (PDP) of the modified S-FED-based NOR, NAND, XOR gates were found to be about 416fJ, 408fJ and 336fJ, respectively, compared with 906fJ, 810fJ and 705fJ achievable with conventional S-FED logic gates.

In this position paper, I argue that a fruitful, and as yet largely unexplored, avenue for artificial life research lies in modelling organisms (specifically, phenotypes) and environment as a single dynamical system. From this perspective, the origin and evolution of life is the progressive control of the dynamical system at a local level by constraints which are represented on an organism's genome. Such an approach shifts the focus of artificial life models away from the design of individuals, towards the interaction of an individual with its dynamic environment. It also blurs the boundary between organism and environment; the most important modelling distinction is no longer between an organism's body and its external environment, but rather between the genome (which is treated as an essentially symbolic structure) and phenotype-plus-environment combined. An evolutionary cellular automata system, called EvoCA, is introduced as a tool to explore these ideas. To demonstrate how this approach differs from traditional studies, two example applications of EvoCA are presented. One concerns sensor and effector evolution; the other concerns the origin of life, and in particular the evolution of genome-regulated self-stabilising dynamics. Advantages of the new approach are summarised, and some potential criticisms are considered. The paper concludes with a discussion of some implications of this shift in perspective.

This thesis is focused on using genetic programming to evolve images based on lightweight features extracted from a given target image. The main motivation of this thesis is research by Lombardi et al. in which an image retrieval system is developed based on lightweight statistical features of images for comparing and classifying them in painting style categories; primarily based on color matching. In this thesis, automatic fitness scoring of variations of up to 17 lightweight image features using many-objective fitness evaluation was used to evolve textures. Evolved results were shown to have similar color characteristics to target images. Although a human survey was conducted to confirm those results, it was inconclusive. Acknowledgements I would like to thank Dr. Brian J. Ross for guiding me in this path patiently and then I would like to thank my parents and my sister, my friends, and faculty of department of computer science for their great support.

SS-CCEA results on LeadingOnes problem with constant mutation. .. . 4.3 SS-CCEA results on s • LeadingOnes − OneMax problem. .. .. .. .. 4.4 Linkage bias results of SS-CCEA on s • LeadingOnes − OneMax. .. .. 4.5 Linkage bias results for the SS-CCEA on CLOB 2,2. .. .. .. .. .. .. . 4.6 Collaboration credit assignment results for SS-CCEA on CLOB 2,2. .. . .

The Building Block Hypothesis (BBH) states that adaptive systems combine good partial solutions (so-called building blocks) to find increasingly better solutions. It is thought that Genetic Algorithms (GAs) implement the BBH. However, for GAs building blocks are semi-theoretical objects in that they are thought only to be implicitly exploited via the selection and crossover operations of a GA. In the current work, we discover a mathematical method to identify the complete set of schemata present in a given population of a GA; as such a natural way to study schema processing (and thus the BBH) is revealed. We demonstrate how this approach can be used both theoretically and experimentally. Theoretically, we show that the search space for good schemata is a complete lattice and that each generation samples a complete sub-lattice of this search space. In addition, we show that combining schemata can only explore a subset of the search space. Experimentally, we compare how well different crossover methods combine building blocks. We find that for most crossover methods approximately 25-35% of building blocks in a generation result from the combination of the previous generation's building blocks. We also find that an increase in the combination of building blocks does not lead to an increase in the efficiency of a GA. To complement this article, we introduce an open source Python package called schematax, which allows one to calculate the schemata present in a population using the methods described in this article.

W e describe the method in which a visually guided swing motion for a 16DOF two-armed bipedal robot is ac uired by applying GA (genetic algorithm) to a NN Jeural network) controller. The evolutionary approach to the acquisition of various motions for robots has been successfully used by many researchers, but most studies have been carried out only through computer simulations. In this research, we adopt a real robot with a complicated body used in a noisy environment. The evolutionary processes are examined an a virtual world constructed on a CRS-CS6400 parallel computer which simulates such factors as swing dynamics, visual processes, noise reduction processes, and timelags in a control system. It took about 2 hours for an artificial evolution to create a successful individual after 50 generations from an initial population of 200 unsuccessful genes. Using the NN decoded from the most successful individual of the last generatation, a real two-armed bipedal robot that could swing successfully was obtained.

We are interested in the role of restricted mating schemes in the context of evolutionary multi-objective algorithms. In this paper, we propose an adaptive assortative mating scheme that uses similarity in the decision space (genotypic assortative mating) and adapts the mating pressure as the search progresses. We show that this mechanism improves the performance of the simple evolutionary algorithm for multi-objective optimisation (SEAMO2) on the multiple knapsack problem.

The diagnosis of diseases can be formulated as a classification problem, making it an NP-hard problem. This is the case for the two problems that this work aims to solve: the classification of tumor samples from patients suspected of having breast cancer as benign or malignant and the classification of samples from patients suspected of having type II diabetes as negative or positive. In order to make accurate diagnoses (classification) of these disorders, our idea is to construct approximate algorithms based on multilayer perceptrons, genetic algorithms, and algorithms that hybridize these alternatives. Numerical experiments enable assessing and contrasting the effectiveness of various approaches using actual data sets. The results demonstrate that, in addition to outperforming algorithms suggested in the literature in terms of performance, our ideas produce outcomes with classification errors that are nearly zero.

Birds demonstrate that flapping-wing flight (FWF) is a versatile flight mode, compatible with hovering, forward flight and gliding to save energy. This extended flight domain would be especially useful on mini-UAVs. However, design is challenging because aerodynamic efficiency is conditioned by complex movements of the wings, and because many interactions exist between morphological (wing area, aspect ratio) and kinematic parameters (flapping frequency, stroke amplitude, wing unfolding). Here we used artificial evolution to optimize these morpho-kinematic features on a simulated 1 kg UAV, equipped with wings articulated at the shoulder and wrist. Flight tests were conducted in a dedicated steady aerodynamics simulator. Parameters generating horizontal flight for minimal mechanical power were retained. Results showed that flight at medium speed (10-12 m s −1) can be obtained for reasonable mechanical power (20 W kg −1), while flight at higher speed (16-20 m s −1) implied increased power (30-50 W kg −1). Flight at low speed (6-8 m s −1) necessitated unrealistic power levels (70-500 W kg −1), probably because our simulator neglected unsteady aerodynamics. The underlying adaptation of morphology and kinematics to varying flight speed were compared to available biological data on the flight of birds.

In this paper, we propose a Multi-Agent based Hyper-Heuristic algorithm for the Winner Determination Problem named MAH 2-WDP. This algorithm explores a set of cooperating agents to select the appropriate operation using learning techniques. MAH 2-WDP is specialized for local search methods and evolutionary methods where the following agents are seeking to improve the search within reinforcement learning: the mediator agent, two local search agents, the perturbation agent and two recombination agents. Our computational study shows that the proposed algorithm performs well on the tested benchmark instances in terms of solution quality.

While navigating their environments it is essential for autonomous mobile robots to actively avoid collisions with obstacles. Flying insects perform this behavioural task with ease relying mainly on information the visual system provides. Here we implement a bioinspired collision avoidance algorithm based on the extraction of nearness information from visual motion on the hexapod walking robot platform HECTOR. The algorithm allows HECTOR to navigate cluttered environments while actively avoiding obstacles.

In this work we propose the hybridization of two techniques to improve the cooperation among the fuzzy rules: the use of rule weights and the Cooperative Rules learning methodology. To do that, the said methodology is extended to include the learning of rule weights within the rule cooperation paradigm. Considering these kinds of techniques could result in important improvements of the system accuracy, maintaining the interpretability to an acceptable level. Keywords: Fuzzy rule-based systems, linguistic ...