Pub Date : 2021-10-01DOI: 10.1007/s10710-021-09411-x
André de Almeida Farzat, Márcio de Oliveira Barros
{"title":"Automatic generation of regular expressions for the Regex Golf challenge using a local search algorithm","authors":"André de Almeida Farzat, Márcio de Oliveira Barros","doi":"10.1007/s10710-021-09411-x","DOIUrl":"https://doi.org/10.1007/s10710-021-09411-x","url":null,"abstract":"","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"135 3-4 1","pages":"105 - 131"},"PeriodicalIF":2.6,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77212279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-09-30DOI: 10.1007/s10710-021-09413-9
Françoso Dal Piccol Sotto, Léo, Kaufmann, Paul, Atkinson, Timothy, Kalkreuth, Roman, Porto Basgalupp, Márcio
Graph representations promise several desirable properties for genetic programming (GP); multiple-output programs, natural representations of code reuse and, in many cases, an innate mechanism for neutral drift. Each graph GP technique provides a program representation, genetic operators and overarching evolutionary algorithm. This makes it difficult to identify the individual causes of empirical differences, both between these methods and in comparison to traditional GP. In this work, we empirically study the behaviour of Cartesian genetic programming (CGP), linear genetic programming (LGP), evolving graphs by graph programming and traditional GP. By fixing some aspects of the configurations, we study the performance of each graph GP method and GP in combination with three different EAs: generational, steady-state and ((1+lambda )). In general, we find that the best choice of representation, genetic operator and evolutionary algorithm depends on the problem domain. Further, we find that graph GP methods can increase search performance on complex real-world regression problems and, particularly in combination with the ((1 + lambda)) EA, are significantly better on digital circuit synthesis tasks. We further show that the reuse of intermediate results by tuning LGP’s number of registers and CGP’s levels back parameter is of utmost importance and contributes significantly to better convergence of an optimization algorithm when solving complex problems that benefit from code reuse.
{"title":"Graph representations in genetic programming","authors":"Françoso Dal Piccol Sotto, Léo, Kaufmann, Paul, Atkinson, Timothy, Kalkreuth, Roman, Porto Basgalupp, Márcio","doi":"10.1007/s10710-021-09413-9","DOIUrl":"https://doi.org/10.1007/s10710-021-09413-9","url":null,"abstract":"<p>Graph representations promise several desirable properties for genetic programming (GP); multiple-output programs, natural representations of code reuse and, in many cases, an innate mechanism for neutral drift. Each graph GP technique provides a program representation, genetic operators and overarching evolutionary algorithm. This makes it difficult to identify the individual causes of empirical differences, both between these methods and in comparison to traditional GP. In this work, we empirically study the behaviour of Cartesian genetic programming (CGP), linear genetic programming (LGP), evolving graphs by graph programming and traditional GP. By fixing some aspects of the configurations, we study the performance of each graph GP method and GP in combination with three different EAs: generational, steady-state and <span>((1+lambda ))</span>. In general, we find that the best choice of representation, genetic operator and evolutionary algorithm depends on the problem domain. Further, we find that graph GP methods can increase search performance on complex real-world regression problems and, particularly in combination with the (<span>(1 + lambda)</span>) EA, are significantly better on digital circuit synthesis tasks. We further show that the reuse of intermediate results by tuning LGP’s number of registers and CGP’s levels back parameter is of utmost importance and contributes significantly to better convergence of an optimization algorithm when solving complex problems that benefit from code reuse.</p>","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"30 3","pages":""},"PeriodicalIF":2.6,"publicationDate":"2021-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138496050","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-09-30DOI: 10.1007/s10710-021-09417-5
A. Saini, L. Spector
{"title":"Relationships between parent selection methods, looping constructs, and success rate in genetic programming","authors":"A. Saini, L. Spector","doi":"10.1007/s10710-021-09417-5","DOIUrl":"https://doi.org/10.1007/s10710-021-09417-5","url":null,"abstract":"","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"22 1","pages":"495 - 509"},"PeriodicalIF":2.6,"publicationDate":"2021-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49407753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-09-03DOI: 10.1007/s10710-021-09412-w
J. Schmitt, S. Kuckuk, H. Köstler
{"title":"EvoStencils: a grammar-based genetic programming approach for constructing efficient geometric multigrid methods","authors":"J. Schmitt, S. Kuckuk, H. Köstler","doi":"10.1007/s10710-021-09412-w","DOIUrl":"https://doi.org/10.1007/s10710-021-09412-w","url":null,"abstract":"","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"22 1","pages":"511 - 537"},"PeriodicalIF":2.6,"publicationDate":"2021-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45676336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study both genotypic and phenotypic convergence in GP floating point continuous domain symbolic regression over thousands of generations. Subtree fitness variation across the population is measured and shown in many cases to fall. In an expanding region about the root node, both genetic opcodes and function evaluation values are identical or nearly identical. Bottom up (leaf to root) analysis shows both syntactic and semantic (including entropy) similarity expand from the outermost node. Despite large regions of zero variation, fitness continues to evolve and near zero crossover disruption suggests improved GP systems within existing memory use.
{"title":"Genetic programming convergence","authors":"W. Langdon","doi":"10.1145/3520304.3534063","DOIUrl":"https://doi.org/10.1145/3520304.3534063","url":null,"abstract":"We study both genotypic and phenotypic convergence in GP floating point continuous domain symbolic regression over thousands of generations. Subtree fitness variation across the population is measured and shown in many cases to fall. In an expanding region about the root node, both genetic opcodes and function evaluation values are identical or nearly identical. Bottom up (leaf to root) analysis shows both syntactic and semantic (including entropy) similarity expand from the outermost node. Despite large regions of zero variation, fitness continues to evolve and near zero crossover disruption suggests improved GP systems within existing memory use.","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"23 1","pages":"71-104"},"PeriodicalIF":2.6,"publicationDate":"2021-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41964547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-08-19DOI: 10.1007/s10710-021-09410-y
Rockett, Peter
This paper extends the numerical tuning of tree constants in genetic programming (GP) to the multiobjective domain. Using ten real-world benchmark regression datasets and employing Bayesian comparison procedures, we first consider the effects of feature standardization (without constant tuning) and conclude that standardization generally produces lower test errors, but, contrary to other recently published work, we find much less clear trend for tree sizes. In addition, we consider the effects of constant tuning – with and without feature standardization – and observe that (1) constant tuning invariably improves test error, and (2) usually decreases tree size. Combined with standardization, constant tuning produces the best test error results; tree sizes, however, are increased. We also examine the effects of applying constant tuning only once at the end a conventional GP run which turns out to be surprisingly promising. Finally, we consider the merits of using numerical procedures to tune tree constants and observe that for around half the datasets evolutionary search alone is superior whereas for the remaining half, parameter tuning is superior. We identify a number of open research questions that arise from this work.
{"title":"Constant optimization and feature standardization in multiobjective genetic programming","authors":"Rockett, Peter","doi":"10.1007/s10710-021-09410-y","DOIUrl":"https://doi.org/10.1007/s10710-021-09410-y","url":null,"abstract":"<p>This paper extends the numerical tuning of tree constants in genetic programming (GP) to the multiobjective domain. Using ten real-world benchmark regression datasets and employing Bayesian comparison procedures, we first consider the effects of feature standardization (without constant tuning) and conclude that standardization generally produces lower test errors, but, contrary to other recently published work, we find much less clear trend for tree sizes. In addition, we consider the effects of constant tuning – with and without feature standardization – and observe that (1) constant tuning invariably improves test error, and (2) usually decreases tree size. Combined with standardization, constant tuning produces the best test error results; tree sizes, however, are increased. We also examine the effects of applying constant tuning only once at the end a conventional GP run which turns out to be surprisingly promising. Finally, we consider the merits of using numerical procedures to tune tree constants and observe that for around half the datasets evolutionary search alone is superior whereas for the remaining half, parameter tuning is superior. We identify a number of open research questions that arise from this work.</p>","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"30 6","pages":""},"PeriodicalIF":2.6,"publicationDate":"2021-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138496049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-08-10DOI: 10.1007/s10710-023-09448-0
M. Moreno, Alexander Lalejini, C. Ofria
{"title":"Matchmaker, matchmaker, make me a match: geometric, variational, and evolutionary implications of criteria for tag affinity","authors":"M. Moreno, Alexander Lalejini, C. Ofria","doi":"10.1007/s10710-023-09448-0","DOIUrl":"https://doi.org/10.1007/s10710-023-09448-0","url":null,"abstract":"","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"24 1","pages":"1-42"},"PeriodicalIF":2.6,"publicationDate":"2021-08-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49475616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-21DOI: 10.1007/s10710-021-09408-6
Danil Koryakin, S. Otte, Martin Volker Butz
{"title":"Inference of time series components by online co-evolution","authors":"Danil Koryakin, S. Otte, Martin Volker Butz","doi":"10.1007/s10710-021-09408-6","DOIUrl":"https://doi.org/10.1007/s10710-021-09408-6","url":null,"abstract":"","PeriodicalId":50424,"journal":{"name":"Genetic Programming and Evolvable Machines","volume":"23 1","pages":"7 - 35"},"PeriodicalIF":2.6,"publicationDate":"2021-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1007/s10710-021-09408-6","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"52297886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号