In this paper we introduce a derivative-free optimization method that is derived from a population based stochastic gradient estimator. We first demonstrate some properties of this estimator and show how it is expected to always yield a descent direction. We analytically show that the difference between the expected function value and the optimum decreases exponentially for strongly convex functions and the expected distance between the current point and the optimum has an upper bound. Then we experimentally tune the parameters of our algorithm to get the best performance. Finally, we use the Black-Box-Optimization-Benchmarking test function suite to evaluate the performance of the algorithm. The experiments indicate that the method offer notable performance advantages especially, when applied to objective functions that are ill-conditioned and potentially multi-modal. This result, coupled with the low computational cost when compared to Quasi-Newton methods, makes it quite attractive.
{"title":"Derivative free optimization using a population-based stochastic gradient estimator","authors":"Azhar Khayrattee, G. Anagnostopoulos","doi":"10.1145/2576768.2598365","DOIUrl":"https://doi.org/10.1145/2576768.2598365","url":null,"abstract":"In this paper we introduce a derivative-free optimization method that is derived from a population based stochastic gradient estimator. We first demonstrate some properties of this estimator and show how it is expected to always yield a descent direction. We analytically show that the difference between the expected function value and the optimum decreases exponentially for strongly convex functions and the expected distance between the current point and the optimum has an upper bound. Then we experimentally tune the parameters of our algorithm to get the best performance. Finally, we use the Black-Box-Optimization-Benchmarking test function suite to evaluate the performance of the algorithm. The experiments indicate that the method offer notable performance advantages especially, when applied to objective functions that are ill-conditioned and potentially multi-modal. This result, coupled with the low computational cost when compared to Quasi-Newton methods, makes it quite attractive.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134355513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Purshouse, Abdallah K. Ally, A. Brennan, Daniel Moyo, P. Norman
This paper presents a new real-world application of evolutionary computation: identifying parameterisations of a theory-driven model that can reproduce alcohol consumption dynamics observed in a population over time. Population alcohol consumption is a complex system, with multiple interactions between economic and social factors and drinking behaviours, the nature and importance of which are not well-understood. Prediction of time trends in consumption is therefore difficult, but essential for robust estimation of future changes in health-related consequences of drinking and for appraising the impact of interventions aimed at changing alcohol use in society. The paper describes a microsimulation approach in which an attitude-behaviour model, Theory of Planned Behaviour, is used to describe the frequency of drinking by individuals. Consumption dynamics in the simulation are driven by changes in the social roles of individuals over time (parenthood, partnership, and paid labour). An evolutionary optimizer is used to identify parameterisations of the Theory that can describe the observed changes in drinking frequency. Niching is incorporated to enable multiple possible parameterisations to be identified, each of which can accurately recreate history but potentially encode quite different future trends. The approach is demonstrated using evidence from the 1979-1985 birth cohort in England between 2003 and 2010.
{"title":"Evolutionary parameter estimation for a theory of planned behaviour microsimulation of alcohol consumption dynamics in an English birth cohort 2003 to 2010","authors":"R. Purshouse, Abdallah K. Ally, A. Brennan, Daniel Moyo, P. Norman","doi":"10.1145/2576768.2598239","DOIUrl":"https://doi.org/10.1145/2576768.2598239","url":null,"abstract":"This paper presents a new real-world application of evolutionary computation: identifying parameterisations of a theory-driven model that can reproduce alcohol consumption dynamics observed in a population over time. Population alcohol consumption is a complex system, with multiple interactions between economic and social factors and drinking behaviours, the nature and importance of which are not well-understood. Prediction of time trends in consumption is therefore difficult, but essential for robust estimation of future changes in health-related consequences of drinking and for appraising the impact of interventions aimed at changing alcohol use in society. The paper describes a microsimulation approach in which an attitude-behaviour model, Theory of Planned Behaviour, is used to describe the frequency of drinking by individuals. Consumption dynamics in the simulation are driven by changes in the social roles of individuals over time (parenthood, partnership, and paid labour). An evolutionary optimizer is used to identify parameterisations of the Theory that can describe the observed changes in drinking frequency. Niching is incorporated to enable multiple possible parameterisations to be identified, each of which can accurately recreate history but potentially encode quite different future trends. The approach is demonstrated using evidence from the 1979-1985 birth cohort in England between 2003 and 2010.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"114 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132782537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Friedrich Burkhard von der Osten, M. Kirley, Tim Miller
Reactive path planning to avoid collisions with moving obstacles enables more robust agent systems. However, many solutions assume that moving objects are passive; that is, they do not consider that the moving objects are themselves re-planning to avoid collisions, and thus may change their trajectory. In this paper we present a model, Anticipatory Stigmergic Collision Avoidance (ASCA) for reciprocal collision avoidance using anticipatory stigmergy. Unlike standard stigmergy, in which agents leave pheromones to indicate a trace of previous actions, anticipatory stigmergy deposits pheromones on intended future paths. By sharing their intended future paths with each other at regular intervals, agents can re-plan to attempt to avoid collisions. We experimentally evaluate ASCA over three scenarios, and compare with a state of art approach, Reciprocal Velocity Obstacles (RVO). Our evaluation showed that ASCA is consistently more robust in noisy environments in which transmitted information can be lost or degraded. Further, using ASCA without noise results in fewer collisions than RVO when agents are in formation, but more collisions when formed randomly.
{"title":"Anticipatory stigmergic collision avoidance under noise","authors":"Friedrich Burkhard von der Osten, M. Kirley, Tim Miller","doi":"10.1145/2576768.2598389","DOIUrl":"https://doi.org/10.1145/2576768.2598389","url":null,"abstract":"Reactive path planning to avoid collisions with moving obstacles enables more robust agent systems. However, many solutions assume that moving objects are passive; that is, they do not consider that the moving objects are themselves re-planning to avoid collisions, and thus may change their trajectory. In this paper we present a model, Anticipatory Stigmergic Collision Avoidance (ASCA) for reciprocal collision avoidance using anticipatory stigmergy. Unlike standard stigmergy, in which agents leave pheromones to indicate a trace of previous actions, anticipatory stigmergy deposits pheromones on intended future paths. By sharing their intended future paths with each other at regular intervals, agents can re-plan to attempt to avoid collisions. We experimentally evaluate ASCA over three scenarios, and compare with a state of art approach, Reciprocal Velocity Obstacles (RVO). Our evaluation showed that ASCA is consistently more robust in noisy environments in which transmitted information can be lost or degraded. Further, using ASCA without noise results in fewer collisions than RVO when agents are in formation, but more collisions when formed randomly.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134324315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this work we demonstrate novel applications of generative grammar to evolutionary search. We introduce a class of grammar that can represent hierarchical schema structure in a problem space, and describe an algorithm that can infer an instance of the grammar from a population of sample phenotypes. Unlike conventional sequence-based grammars this grammar represents set-membership relationships, not strings, and is therefore insensitive to gene-ordering and physical linkage. We show that these methods are capable of accurately identifying problem structure from populations of above-average-fitness individuals on simple modular and hierarchically modular test problems. We then show how these grammatical models can be used to aid evolutionary problem solving by enabling facilitated variation; specifically, by producing novel combinations of schemata observed in the sample population whilst respecting the inherent constraint structure of the problem space. This provides a robust method of building-block recombination that is linkage-invariant and not restricted to low-order schemata.
{"title":"Solving building block problems using generative grammar","authors":"Chris R. Cox, R. Watson","doi":"10.1145/2576768.2598259","DOIUrl":"https://doi.org/10.1145/2576768.2598259","url":null,"abstract":"In this work we demonstrate novel applications of generative grammar to evolutionary search. We introduce a class of grammar that can represent hierarchical schema structure in a problem space, and describe an algorithm that can infer an instance of the grammar from a population of sample phenotypes. Unlike conventional sequence-based grammars this grammar represents set-membership relationships, not strings, and is therefore insensitive to gene-ordering and physical linkage. We show that these methods are capable of accurately identifying problem structure from populations of above-average-fitness individuals on simple modular and hierarchically modular test problems. We then show how these grammatical models can be used to aid evolutionary problem solving by enabling facilitated variation; specifically, by producing novel combinations of schemata observed in the sample population whilst respecting the inherent constraint structure of the problem space. This provides a robust method of building-block recombination that is linkage-invariant and not restricted to low-order schemata.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"8 6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116799401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jeremy Acre, B. E. Eskridge, Nicholas Zoller, I. Schlupp
Many animals form large aggregations that have no apparent consistent leader, yet are capable of highly coordinated movements. At any given time, it seems like an individual can emerge as a leader only to be replaced by another. Although individuals within a group are largely considered equal, even individuals in a homogeneous group are different. Clearly individuals will differ based on traits like sex, age, and experience. Of particular interest is the idea of individuals differing in their correlated traits, or personality. Different personalities can arise via complex interactions between genes and an environment and are often shaped by individual experience. For example, one would generally predict that individuals characterized as "bold" would more frequently be leaders. However, if the environment changes, how do once successful leaders respond to failure and how do newly successful leaders emerge? Using a biologically-based collective movement model, we demonstrate that a self-assessment mechanism using winner and loser effects is capable of producing transitory leaders who change roles in response to changes in the environment. Furthermore, simulations predict that this self-assessment mechanism allows the group to adapt to drastic changes in the environment and remain successful.
{"title":"Adapting to a changing environment using winner and loser effects","authors":"Jeremy Acre, B. E. Eskridge, Nicholas Zoller, I. Schlupp","doi":"10.1145/2576768.2598355","DOIUrl":"https://doi.org/10.1145/2576768.2598355","url":null,"abstract":"Many animals form large aggregations that have no apparent consistent leader, yet are capable of highly coordinated movements. At any given time, it seems like an individual can emerge as a leader only to be replaced by another. Although individuals within a group are largely considered equal, even individuals in a homogeneous group are different. Clearly individuals will differ based on traits like sex, age, and experience. Of particular interest is the idea of individuals differing in their correlated traits, or personality. Different personalities can arise via complex interactions between genes and an environment and are often shaped by individual experience. For example, one would generally predict that individuals characterized as \"bold\" would more frequently be leaders. However, if the environment changes, how do once successful leaders respond to failure and how do newly successful leaders emerge? Using a biologically-based collective movement model, we demonstrate that a self-assessment mechanism using winner and loser effects is capable of producing transitory leaders who change roles in response to changes in the environment. Furthermore, simulations predict that this self-assessment mechanism allows the group to adapt to drastic changes in the environment and remain successful.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"371 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121742955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dealing with high-dimensional input spaces, like visual input, is a challenging task for reinforcement learning (RL). Neuroevolution (NE), used for continuous RL problems, has to either reduce the problem dimensionality by (1) compressing the representation of the neural network controllers or (2) employing a pre-processor (compressor) that transforms the high-dimensional raw inputs into low-dimensional features. In this paper, we are able to evolve extremely small recurrent neural network (RNN) controllers for a task that previously required networks with over a million weights. The high-dimensional visual input, which the controller would normally receive, is first transformed into a compact feature vector through a deep, max-pooling convolutional neural network (MPCNN). Both the MPCNN preprocessor and the RNN controller are evolved successfully to control a car in the TORCS racing simulator using only visual input. This is the first use of deep learning in the context evolutionary RL.
{"title":"Evolving deep unsupervised convolutional networks for vision-based reinforcement learning","authors":"J. Koutník, J. Schmidhuber, F. Gomez","doi":"10.1145/2576768.2598358","DOIUrl":"https://doi.org/10.1145/2576768.2598358","url":null,"abstract":"Dealing with high-dimensional input spaces, like visual input, is a challenging task for reinforcement learning (RL). Neuroevolution (NE), used for continuous RL problems, has to either reduce the problem dimensionality by (1) compressing the representation of the neural network controllers or (2) employing a pre-processor (compressor) that transforms the high-dimensional raw inputs into low-dimensional features. In this paper, we are able to evolve extremely small recurrent neural network (RNN) controllers for a task that previously required networks with over a million weights. The high-dimensional visual input, which the controller would normally receive, is first transformed into a compact feature vector through a deep, max-pooling convolutional neural network (MPCNN). Both the MPCNN preprocessor and the RNN controller are evolved successfully to control a car in the TORCS racing simulator using only visual input. This is the first use of deep learning in the context evolutionary RL.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123460963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. Hidalgo, J. Colmenar, J. L. Risco-Martín, Carlos Sánchez-Lacruz, J. Lanchares, O. Garnica, Josefa Díaz
Different techniques have been proposed to tackle GA-Hard problems. Some techniques work with different encodings and representations, other use reordering operators and several, such as the Evolutionary Mapping Method (EMM), apply genotype-phenotype mappings. EMM uses multiple chromosomes in a single cell for mating with another cell within a single population. Although EMM gave good results, it fails on solving some deceptive problems. In this line, EMMRS (EMM with Replacement and Shift) adds a new operator, consisting on doing a replacement and a shift of some of the bits within the chromosome. Results showed the efficiency of the proposal on deceptive problems. However, EMMRS was not tested with other kind of hard problems. In this paper we have adapted EMMRS for solving the Traveling Salesman Problem (TSP). The encodings and genetic operators for solving the TSP are quite different to those applied on deceptive problems. In addition, execution times recommended the parallelization of the GA. We implemented a GPU parallel version. We present here some preliminary results proving that Evolutionary Mapping Method with Replacement and Shift gives good results not only in terms of quality but also in terms of speedup on its GPU parallel version for some instances of the TSP problem.
{"title":"Solving GA-hard problems with EMMRS and GPGPUs","authors":"J. Hidalgo, J. Colmenar, J. L. Risco-Martín, Carlos Sánchez-Lacruz, J. Lanchares, O. Garnica, Josefa Díaz","doi":"10.1145/2576768.2598219","DOIUrl":"https://doi.org/10.1145/2576768.2598219","url":null,"abstract":"Different techniques have been proposed to tackle GA-Hard problems. Some techniques work with different encodings and representations, other use reordering operators and several, such as the Evolutionary Mapping Method (EMM), apply genotype-phenotype mappings. EMM uses multiple chromosomes in a single cell for mating with another cell within a single population. Although EMM gave good results, it fails on solving some deceptive problems. In this line, EMMRS (EMM with Replacement and Shift) adds a new operator, consisting on doing a replacement and a shift of some of the bits within the chromosome. Results showed the efficiency of the proposal on deceptive problems. However, EMMRS was not tested with other kind of hard problems. In this paper we have adapted EMMRS for solving the Traveling Salesman Problem (TSP). The encodings and genetic operators for solving the TSP are quite different to those applied on deceptive problems. In addition, execution times recommended the parallelization of the GA. We implemented a GPU parallel version. We present here some preliminary results proving that Evolutionary Mapping Method with Replacement and Shift gives good results not only in terms of quality but also in terms of speedup on its GPU parallel version for some instances of the TSP problem.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125472250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A variety of real world applications fit into the broad definition of time series classification. Using traditional machine learning approaches such as treating the time series sequences as high dimensional vectors have faced the well known "curse of dimensionality" problem. Recently, the field of time series classification has seen success by using preprocessing steps that discretize the time series using a Symbolic Aggregate ApproXimation technique (SAX) and using recurring subsequences ("motifs") as features. In this paper we explore a feature construction algorithm based on genetic programming that uses SAX-generated motifs as the building blocks for the construction of more complex features. The research shows that the constructed complex features improve the classification accuracy in a statistically significant manner for many applications.
{"title":"SAX-EFG: an evolutionary feature generation framework for time series classification","authors":"Uday Kamath, Jessica Lin, K. D. Jong","doi":"10.1145/2576768.2598321","DOIUrl":"https://doi.org/10.1145/2576768.2598321","url":null,"abstract":"A variety of real world applications fit into the broad definition of time series classification. Using traditional machine learning approaches such as treating the time series sequences as high dimensional vectors have faced the well known \"curse of dimensionality\" problem. Recently, the field of time series classification has seen success by using preprocessing steps that discretize the time series using a Symbolic Aggregate ApproXimation technique (SAX) and using recurring subsequences (\"motifs\") as features. In this paper we explore a feature construction algorithm based on genetic programming that uses SAX-generated motifs as the building blocks for the construction of more complex features. The research shows that the constructed complex features improve the classification accuracy in a statistically significant manner for many applications.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129085450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we use Differential Evolution (DE), with best evolved results refined using a Nelder-Mead optimization, to solve complex problems in orbital mechanics relevant to low Earth orbits (LEO). A class of so-called 'Lambert Problems' is examined. We evolve impulsive initial velocity vectors giving rise to intercept trajectories that take a spacecraft from given initial positions to specified target positions. We seek to minimize final positional error subject to time-of-flight and/or energy (fuel) constraints. We first validate that the method can recover known analytical solutions obtainable with the assumption of Keplerian motion. We then apply the method to more complex and realistic non-Keplerian problems incorporating trajectory perturbations arising in LEO due to the Earth's oblateness and rarefied atmospheric drag. The viable trajectories obtained for these difficult problems suggest the robustness of our computational approach for real-world orbital trajectory design in LEO situations where no analytical solution exists.
{"title":"Evolved spacecraft trajectories for low earth orbit","authors":"D. Hinckley, Karol Zieba, D. Hitt, M. Eppstein","doi":"10.1145/2576768.2598246","DOIUrl":"https://doi.org/10.1145/2576768.2598246","url":null,"abstract":"In this paper we use Differential Evolution (DE), with best evolved results refined using a Nelder-Mead optimization, to solve complex problems in orbital mechanics relevant to low Earth orbits (LEO). A class of so-called 'Lambert Problems' is examined. We evolve impulsive initial velocity vectors giving rise to intercept trajectories that take a spacecraft from given initial positions to specified target positions. We seek to minimize final positional error subject to time-of-flight and/or energy (fuel) constraints. We first validate that the method can recover known analytical solutions obtainable with the assumption of Keplerian motion. We then apply the method to more complex and realistic non-Keplerian problems incorporating trajectory perturbations arising in LEO due to the Earth's oblateness and rarefied atmospheric drag. The viable trajectories obtained for these difficult problems suggest the robustness of our computational approach for real-world orbital trajectory design in LEO situations where no analytical solution exists.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129161564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We propose a new means of executing a genetic program which improves its output quality. Our approach, called Multiple Regression Genetic Programming (MRGP) decouples and linearly combines a program's subexpressions via multiple regression on the target variable. The regression yields an alternate output: the prediction of the resulting multiple regression model. It is this output, over many fitness cases, that we assess for fitness, rather than the program's execution output. MRGP can be used to improve the fitness of a final evolved solution. On our experimental suite, MRGP consistently generated solutions fitter than the result of competent GP or multiple regression. When integrated into GP, inline MRGP, on the basis of equivalent computational budget, outperforms competent GP while also besting post-run MRGP. Thus MRGP's output method is shown to be superior to the output of program execution and it represents a practical, cost neutral, improvement to GP.
{"title":"Multiple regression genetic programming","authors":"Ignacio Arnaldo, K. Krawiec, Una-May O’Reilly","doi":"10.1145/2576768.2598291","DOIUrl":"https://doi.org/10.1145/2576768.2598291","url":null,"abstract":"We propose a new means of executing a genetic program which improves its output quality. Our approach, called Multiple Regression Genetic Programming (MRGP) decouples and linearly combines a program's subexpressions via multiple regression on the target variable. The regression yields an alternate output: the prediction of the resulting multiple regression model. It is this output, over many fitness cases, that we assess for fitness, rather than the program's execution output. MRGP can be used to improve the fitness of a final evolved solution. On our experimental suite, MRGP consistently generated solutions fitter than the result of competent GP or multiple regression. When integrated into GP, inline MRGP, on the basis of equivalent computational budget, outperforms competent GP while also besting post-run MRGP. Thus MRGP's output method is shown to be superior to the output of program execution and it represents a practical, cost neutral, improvement to GP.","PeriodicalId":123241,"journal":{"name":"Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"176 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129545965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：