Abstract. It has been suggested that evolving developmental programs instead of direct genotype-p... more Abstract. It has been suggested that evolving developmental programs instead of direct genotype-phenotype mappings may increase the scalability of Genetic Algorithms. Many of these Artificial Embryogeny (AE) models have been proposed and their evolutionary properties are being investigated. One of these properties concerns the fault-tolerance of at least a particular class of AE, which models the development of artificial multicellular organisms. It has been shown that such AE evolves designs capable of recovering phenotypic faults during development, even if faulttolerance is not selected for during evolution. This type of adaptivity is clearly very interesting both for theoretical reasons and possible robotic applications. In this paper we provide empirical evidence collected from a multicellular AE model showing a subtle relationship between evolution and development. These results explain why developmental fault-tolerance necessarily emerges during evolution. 1
With a gene required for each phenotypic trait, direct genetic encodings may show poor scalabilit... more With a gene required for each phenotypic trait, direct genetic encodings may show poor scalability to increasing phenotype length. Developmental systems may alleviate this problem by providing more efficient indirect genotype to phenotype mappings. A novel classification of multi-cellular developmental systems in evolvable hardware is introduced. It shows a category of developmental systems that up to now has rarely been explored. We argue that this category is where most of the benefits of developmental systems lie (e.g. speed, scalability, robustness, inter-cellular and environmental interactions that allow fault-tolerance or adaptivity). This article describes a very simple genetic encoding and developmental system designed for multi-cellular circuits that belongs to this category. We refer to it as the morphogenetic system. The morphogenetic system is inspired by gene expression and cellular differentiation. It focuses on low computational requirements which allows fast executio...
Abstract- It is recognized that the combination of genetic and local search can have strong syner... more Abstract- It is recognized that the combination of genetic and local search can have strong synergistic effeets. In same cases though, the local search mechanism can be too aggressive, mislead the evolutionary search and produce premature convergence. We set up a population of evolving agents also capable of learning by operant conditioning and communicating acquired behaviors (memes). The diffusion and discovery of meme gives rise to a second process of evolution atop of the genetic one. Memes are shown to have both guiding and hiding effects on baldwinian and lamarckian evolution. In contraposition to previous models, simulations show that back-coding of acquired behaviors is highly beneficial only at the beginning of the evolutionary search. This result arises because of the different nature of the guiding provided by memes and the hiding effect that they generate. To minimize the negative influence of the hiding effect but still benefit from the memetic guidance, we decrease the...
The Implicant Network is a neural network model capable of storing an arbitrary boolean function ... more The Implicant Network is a neural network model capable of storing an arbitrary boolean function F:{0,1}^n -> {0,1}. The difference from previous one-shot learning models is that the training algorithm compresses the positive set online with linear time and space requirements. The algorithm works by building a Sum Of Products (SOP) representation of the positive set as it is presented to the network. Since the minimum coverage of implicants is an NP-Hard problem, the compression rate is not optimal at first but it is shown to increase rapidly as the positive set is shown over again.
Evolving large phenotypes remains nowadays a problem due to the combinatorial explosion of the se... more Evolving large phenotypes remains nowadays a problem due to the combinatorial explosion of the search space. Seeking better scalability and inspired by the development of biological systems several indirect genetic encodings have been proposed. Here two different developmental mechanisms are compared. The first, developed for hardware implementations, relies on simple mechanisms inspired upon gene regulation and cell differentiation. The second, inspired by Cellular Automata, is an Artificial Embryogeny system based on cell-chemistry. This paper analyses the scalability and robustness to phenotypic faults of these two systems, with a direct encoding strategy used for comparison. Results show
Multi-level grounding and self-organization of behaviour through evolution, development, learning and culture
The research topic of this PhD concerns the production of intelligent adaptive behaviour from the... more The research topic of this PhD concerns the production of intelligent adaptive behaviour from the embodied perspective.The intelligence seen in biological organisms arises as an emergent property of a complex system displaying multiple levels of adaptive mechanisms. At the bottom, slow changes to the genotype structure are mediated by natural selection. At the top, to cope with more transient factors, faster lifetime transformations are found.Lifetime adaptation also takes place at various levels, from the processes that are responsible for the organism construction and maintenance, to the regulation of behaviour based on instinctive responses, cognitive abilities and social interactions.The traditional GOFAI (Good Old Fashion Artificial Intelligence) stand is based on the view of intelligence as computation taking place on a internal symbolic representations of the outside world. Powered by deductive logic, the classical artificial reasoner is centered on the possibility to describ...
Ascalable evolutionary model for spiking neural networks : development with embryonal stages
The research topic of this PhD concerns the production of intelligent adaptive behaviour from the... more The research topic of this PhD concerns the production of intelligent adaptive behaviour from the embodied perspective.The intelligence seen in biological organisms arises as an emergent property of a complex system displaying multiple levels of adaptive mechanisms. At the bottom, slow changes to the genotype structure are mediated by natural selection. At the top, to cope with more transient factors, faster lifetime transformations are found.Lifetime adaptation also takes place at various levels, from the processes that are responsible for the organism construction and maintenance, to the regulation of behaviour based on instinctive responses, cognitive abilities and social interactions.The traditional GOFAI (Good Old Fashion Artificial Intelligence) stand is based on the view of intelligence as computation taking place on a internal symbolic representations of the outside world. Powered by deductive logic, the classical artificial reasoner is centered on the possibility to describ...
Increasing the evolvability of development with Embryonal Stages
The research topic of this PhD concerns the production of intelligent adaptive behaviour from the... more The research topic of this PhD concerns the production of intelligent adaptive behaviour from the embodied perspective.The intelligence seen in biological organisms arises as an emergent property of a complex system displaying multiple levels of adaptive mechanisms. At the bottom, slow changes to the genotype structure are mediated by natural selection. At the top, to cope with more transient factors, faster lifetime transformations are found.Lifetime adaptation also takes place at various levels, from the processes that are responsible for the organism construction and maintenance, to the regulation of behaviour based on instinctive responses, cognitive abilities and social interactions.The traditional GOFAI (Good Old Fashion Artificial Intelligence) stand is based on the view of intelligence as computation taking place on a internal symbolic representations of the outside world. Powered by deductive logic, the classical artificial reasoner is centered on the possibility to describ...
Indirect encoding methods are aimed at the reduction of the combinatorial explosion of search spa... more Indirect encoding methods are aimed at the reduction of the combinatorial explosion of search spaces, therefore increasing the evolvability of large phenotypes. These so called Artificial Embryogeny systems have so far shown increased scalability for problems involving solutions of low complexity. This leaves open the more general question about the evolvability of complex phenotypes. In this paper, we introduce a novel method of cellular growth regulated by a developmental program. Genotypes are selected for their ability to develop organisms of specific shape and cell types. Results show that the use of Embryonic Stages, which involves the incremental addition of growth programs, displays positive effects on the evolvability of development.
Inspired upon the development of living systems, many models of artificial embryogeny are being p... more Inspired upon the development of living systems, many models of artificial embryogeny are being proposed. These are usually aimed at the solution of some know limitations of evolutionary computation; among these scalability, flexibility and, more recently, fault-tolerance. This paper focuses on the latter, proposing an explanation of the intrinsic regenerative capabilities displayed by some models of multi-cellular development. Supported by the evidence collected from simulations, regeneration is shown to emerge as evolution converges to more regular regions of the genotype space. The conclusion is that intrinsic fault-tolerance emerges as evolution increases the evolvability of the development process.
It is believed that the second phase of the Baldwin effect is basically governed by the cost of l... more It is believed that the second phase of the Baldwin effect is basically governed by the cost of learning. In this paper we argue that when learning takes place the fitness landscape undergoes a modification that might block the Baldwin effect even if the cost of learning is high. The argument is that learning strategies will bias the evolutionary process towards individuals that genetically acquire better compared to individuals that genetically behave better. Once this process starts the probability of experiencing the Baldwin effect decreases dramatically, whatever the learning cost. A simulation with evolving learning individuals capable of communication is set to show this effect. The set of acquired behaviors (culture) competes with the instinctive one (genes) giving rise to a co-evolutionary effect.
We introduce a model of cellular growth that generates neurocontrollers capable of guiding simple... more We introduce a model of cellular growth that generates neurocontrollers capable of guiding simple simulated agents in a harvesting task.
Uploads
Papers by Diego Federici