Results 1 - 10
of
25
Simulating Evolution with a Computational Model of Embryogeny
, 2006
"... Natural evolution is an incredibly complex dynamical system which we are still grasping to understand. One could argue that the evolutionary search process itself is not the most important part of evolution, but rather it is the evolved representational structures and physical implementations of the ..."
Abstract
-
Cited by 16 (1 self)
- Add to MetaCart
Natural evolution is an incredibly complex dynamical system which we are still grasping to understand. One could argue that the evolutionary search process itself is not the most important part of evolution, but rather it is the evolved representational structures and physical implementations of the genotype-phenotype mappings that make it so powerful. One such behaviour is embryogeny, the process by which genetic representations within cells control the development of a multi-cellular organism. Evolution of complex organisms is highly dependent upon biological mechanisms such as embryogeny, and this has fundamental consequences. The objective of this work is to investigate these consequences for a simulated evolutionary search process using a computational model of embryogeny. The model simulates the dynamics achieved within biological systems between the genetic representation of an individual and the mapping to its phenotypic representation. It is not an attempt to produce a biologically accurate or plausible model of the cell, growth or genetics. The model is used to create individuals which develop into multi-component patterns
Improving evolvability through novelty search and self-adaptation
- in Evolutionary Computation (CEC), 2011 IEEE Congress on. IEEE
"... Abstract—A challenge for current evolutionary algorithms is to yield highly evolvable representations like those in nature. Such evolvability in natural evolution is encouraged through selection: Lineages better at molding to new niches are less susceptible to extinction. Similar selection pressure ..."
Abstract
-
Cited by 13 (5 self)
- Add to MetaCart
(Show Context)
Abstract—A challenge for current evolutionary algorithms is to yield highly evolvable representations like those in nature. Such evolvability in natural evolution is encouraged through selection: Lineages better at molding to new niches are less susceptible to extinction. Similar selection pressure is not generally present in evolutionary algorithms; however, the first hypothesis in this paper is that novelty search, a recent evolutionary technique, also selects for evolvability because it rewards lineages able to continually radiate new behaviors. Results in experiments in a maze-navigation domain in this paper support that novelty search finds more evolvable representations than regular fitnessbased search. However, though novelty search outperforms fitnessbased search in a second biped locomotion experiment, it proves no more evolvable than fitness-based search because delicately balanced behaviors are more fragile in that domain. The second hypothesis is that such fragility can be mitigated through selfadaption, whereby genomes influence their own reproduction. Further experiments in fragile domains with novelty search and self-adaption indeed demonstrate increased evolvability, while, interestingly, adding self-adaptation to fitness-based search decreases evolvability. Thus, selecting for novelty may often facilitate evolvability when representations are not overly fragile; furthermore, achieving the potential of self-adaptation may often critically depend upon the reward scheme driving evolution. I.
Local Optima Networks of NK Landscapes with Neutrality
"... In previous work, we have introduced a network-based model that abstracts many details of the underlying landscape and compresses the landscape information into a weighted, oriented graph which we call the local optima network. The vertices of this graph are the local optima of the given fitness lan ..."
Abstract
-
Cited by 13 (9 self)
- Add to MetaCart
In previous work, we have introduced a network-based model that abstracts many details of the underlying landscape and compresses the landscape information into a weighted, oriented graph which we call the local optima network. The vertices of this graph are the local optima of the given fitness landscape, while the arcs are transition probabilities between local optima basins. Here, we extend this formalism to neutral fitness landscapes, which are common in difficult combinatorial search spaces. By using two known neutral variants of the NK family (i.e. NKp and NKq) in which the amount of neutrality can be tuned by a parameter, we show that our new definitions of the optima networks and the associated basins are consistent with the previous definitions for the non-neutral case. Moreover, our empirical study and statistical analysis show that the features of neutral landscapes interpolate smoothly between landscapes with maximum neutrality and non-neutral ones. We found some unknown structural differences between the two studied families of neutral landscapes. But overall, the network features studied confirmed that neutrality, in landscapes with percolating neutral networks, may enhance heuristic search. Our current methodology requires the exhaustive enumeration of the underlying search space. Therefore, sampling techniques should be developed before this analysis can have practical implications. We argue, however, that the proposed model offers a new perspective into the problem difficulty of combinatorial optimization problems and may inspire the design of more effective search heuristics.
Hyper-ellipsoidal Conditions In XCS: Rotation, Linear Approximation, and Solution Structure ABSTRACT
"... The learning classifier system XCS is an iterative rulelearning system that evolves rule structures based on gradient-based prediction and rule quality estimates. Besides classification and reinforcement learning tasks, XCS was applied as an effective function approximator. Hereby, XCS learns space ..."
Abstract
-
Cited by 11 (8 self)
- Add to MetaCart
(Show Context)
The learning classifier system XCS is an iterative rulelearning system that evolves rule structures based on gradient-based prediction and rule quality estimates. Besides classification and reinforcement learning tasks, XCS was applied as an effective function approximator. Hereby, XCS learns space partitions to enable a maximally accurate and general function approximation. Recently, the function approximation approach was improved by replacing (1) hyperrectangular conditions with hyper-ellipsoids and (2) iterative linear approximation with the recursive least squares method. This paper combines the two approaches assessing the usefulness of each. The evolutionary process is further improved by changing the mutation operator implementing an angular mutation that rotates ellipsoidal structures explicitly. Both enhancements improve XCS performance in various non-linear functions. We also analyze the evolving ellipsoidal structures confirming that XCS stretches and rotates the evolving ellipsoids according to the shape of the underlying function. The results confirm that improvements in both the evolutionary approach and the gradient approach can result in significantly better performance. Categories and Subject Descriptors
Enzyme Genetic Programming: Modelling Biological Evolvability in Genetic Programming
, 2003
"... This thesis is dedicated to all my family and friends. (So you’d better read it! — I’ll be asking questions!) 2 This thesis introduces a new approach to program representation in genetic program-ming in which interactions between program components are expressed in terms of a component’s behaviour ..."
Abstract
-
Cited by 7 (4 self)
- Add to MetaCart
This thesis is dedicated to all my family and friends. (So you’d better read it! — I’ll be asking questions!) 2 This thesis introduces a new approach to program representation in genetic program-ming in which interactions between program components are expressed in terms of a component’s behaviour rather through its relative position within a representation or through other non-behavioural systems of reference. This approach has the advantage that a component’s behaviour is expressed in a way that is independent of any par-ticular program it finds itself within; and thereby overcomes the problem when using conventional program representations whereby program components lose their be-havioural context following recombination. More generally, this implicit context rep-resentation leads to a process of meaningful variation filtering; whereby inappropriate change induced by variation operators can be wholly or partially ignored. This occurs as a consequence of program behaviours emerging from the self-organisation of pro-
Selecting for evolvable representations
- In Proc. of the 2006 Genetic and evolutionary computation conference
, 2006
"... Evolutionary algorithms tend to produce solutions that are not evolvable: Although current fitness may be high, further search is impeded as the effects of mutation and crossover become increasingly detrimental. In nature, in addition to having high fitness, organisms have evolvable genomes: phenoty ..."
Abstract
-
Cited by 6 (2 self)
- Add to MetaCart
(Show Context)
Evolutionary algorithms tend to produce solutions that are not evolvable: Although current fitness may be high, further search is impeded as the effects of mutation and crossover become increasingly detrimental. In nature, in addition to having high fitness, organisms have evolvable genomes: phenotypic variation resulting from random mutation is structured and robust. Evolvability is important because it allows the population to produce meaningful variation, leading to efficient search. However, because evolvability does not improve immediate fitness, it must be selected for indirectly. One way to establish such a selection pressure is to change the fitness function systematically. Under such conditions, evolvability emerges only if the representation allows manipulating how genotypic variation maps onto phenotypic variation and if such manipulations lead to detectable changes in fitness. This research forms a framework for understanding how fitness function and representation interact to produce evolvability. Ultimately evolvable encodings may lead to evolutionary algorithms that exhibit the structured complexity and robustness found in nature.
Deceptiveness and Neutrality The ND Family of Fitness Landscapes
"... When a considerable number of mutations have no effects on fitness values, the fitness landscape is said neutral. In order to study the interplay between neutrality, which exists in many real-world applications, and performances of metaheuristics, it is useful to design landscapes which make it poss ..."
Abstract
-
Cited by 4 (0 self)
- Add to MetaCart
(Show Context)
When a considerable number of mutations have no effects on fitness values, the fitness landscape is said neutral. In order to study the interplay between neutrality, which exists in many real-world applications, and performances of metaheuristics, it is useful to design landscapes which make it possible to tune precisely neutral degree distribution. Even though many neutral landscape models have already been designed, none of them are general enough to create landscapes with specific neutral degree distributions. We propose three steps to design such landscapes: first using an algorithm we construct a landscape whose distribution roughly fits the target one, then we use a simulated annealing heuristic to bring closer the two distributions and finally we affect fitness values to each neutral network. Then using this new family of fitness landscapes we are able to highlight the interplay between deceptiveness and neutrality.
Flamm C: A Sequence-to-Function Map for Ribozyme-catalyzed Metabolisms
- ECAL, Lect Notes Comp Sci
, 2009
"... Abstract. We introduce a novel genotype-phenotype mapping based on the relation between RNA sequence and its secondary structure for the use in evolutionary studies. Various extensive studies concerning RNA folding in the context of neutral theory yielded insights about properties of the structure s ..."
Abstract
-
Cited by 4 (1 self)
- Add to MetaCart
(Show Context)
Abstract. We introduce a novel genotype-phenotype mapping based on the relation between RNA sequence and its secondary structure for the use in evolutionary studies. Various extensive studies concerning RNA folding in the context of neutral theory yielded insights about properties of the structure space and the mapping itself. We intend to get a better understanding of some of these properties and especially of the evolution of RNA-molecules as well as their effect on the evolution of the entire molecular system. We investigate the constitution of the neutral network and compare our mapping with other artificial approaches using cellular automatons, random boolean networks and others also based on RNA folding. We yield the highest extent, connectivity and evolvability of the underlying neutral network. Further, we successfully apply the mapping in an existing model for the evolution of a ribozyme-catalyzed metabolism. Key words: genotype-phenotype map, RNA secondary structure, neutral networks, evolution, robustness, evolvability 1
Neutrality and Variability: Two Sides of Evolvability in Linear Genetic Programming ABSTRACT
, 2009
"... The notion of evolvability has been put forward to describe the “core mechanism ” of natural and artificial evolution. Recently, studies have revealed the influence of the environment upon a system’s evolvability. In this contribution, we study the evolvability of a system in various environmental s ..."
Abstract
-
Cited by 3 (2 self)
- Add to MetaCart
The notion of evolvability has been put forward to describe the “core mechanism ” of natural and artificial evolution. Recently, studies have revealed the influence of the environment upon a system’s evolvability. In this contribution, we study the evolvability of a system in various environmental situations. We consider neutrality and variability as two sides of evolvability. The former makes a system tolerant to mutations and provides a hidden staging ground for future phenotypic changes. The latter produces explorative variations yielding phenotypic improvements. Which of the two dominates is influenced by the environment. We adopt two tools for this study of evolvability: i) the rate of adaptive evolution, which captures the observable adaptive variations driven by evolvability; and ii) the variability of individuals, which measures the potential of an individual to vary functionally. We apply these tools to a Linear Genetic Programming system and observe that evolvability is able to exploit its two sides in different environmental situations.
Evolution through the search for novelty
, 2012
"... I present a new approach to evolutionary search called novelty search, wherein only behavioral novelty is rewarded, thereby abstracting evolution as a search for novel forms. This new approach contrasts with the traditional approach of rewarding progress towards the objective through an objective fu ..."
Abstract
-
Cited by 2 (1 self)
- Add to MetaCart
I present a new approach to evolutionary search called novelty search, wherein only behavioral novelty is rewarded, thereby abstracting evolution as a search for novel forms. This new approach contrasts with the traditional approach of rewarding progress towards the objective through an objective function. Although they are designed to light a path to the objective, objective functions can instead deceive search into converging to dead ends called local optima. As a significant problem in evolutionary computation, deception has inspired many techniques designed to mitigate it. However, nearly all such methods are still ultimately susceptible to deceptive local optima because they still measure progress with respect to the objective, which this dissertation will show is often a broken compass. Furthermore, although novelty search completely abandons the objective, it counterintuitively often outperforms methods that search directly for the objective in deceptive tasks and can induce evolutionary dynamics closer in spirit to natural evolution. The main contributions are to
