A Vectorial Approach to Genetic Programming

Irene Azzali, Leonardo Vanneschi, Sara Silva, Illya Bakurov, Mario Giacobini

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Among the various typologies of problems to which Genetic Programming (GP) has been applied since its origins, symbolic regression is one of the most popular. A common situation consists in the prediction of a target time series based on scalar features and other time series variables collected from multiple subjects. To manage this problem with GP data needs a panel representation where each observation corresponds to a collection on a subject at a precise time instant. However, representing data in this form may imply a loss of information: for instance, the algorithm may not be able to recognize observations belonging to the same subject and their recording order. To maintain the source of knowledge supplied by ordered sequences as time series, we propose a new approach to GP that keeps instances of the same observation together in a vector, introducing vectorial variables as terminals. This new representation allows aggregate functions in the primitive GP set, included with the purpose of describing the behaviour of vectorial variables. In this work, we perform a comparative analysis of vectorial GP (VE-GP) against standard GP (ST-GP). Experiments are conducted on different benchmark problems to highlight the advantages of this new approach.

Original languageEnglish
Title of host publicationGenetic Programming
Subtitle of host publication22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings
EditorsNuno Lourenço, Ting Hu, Hendrik Richter, Lukas Sekanina, Pablo García-Sánchez
Place of PublicationSwitzerland
PublisherSpringer Verlag
Pages213-227
Number of pages15
ISBN (Print)9783030166694
DOIs
Publication statusPublished - 1 Jan 2019
Event22nd European Conference on Genetic Programming, EuroGP 2019, held as part of EvoStar 2019 - Leipzig, Germany
Duration: 24 Apr 201926 Apr 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11451 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference22nd European Conference on Genetic Programming, EuroGP 2019, held as part of EvoStar 2019
CountryGermany
CityLeipzig
Period24/04/1926/04/19

Fingerprint

Genetic programming
Genetic Programming
Time series
Symbolic Regression
Comparative Analysis
Instant
Scalar
Benchmark
Imply
Target
Prediction
Experiment
Observation
Experiments

Keywords

  • Genetic programming
  • Panel data regression
  • Vector-based representation

Cite this

Azzali, I., Vanneschi, L., Silva, S., Bakurov, I., & Giacobini, M. (2019). A Vectorial Approach to Genetic Programming. In N. Lourenço, T. Hu, H. Richter, L. Sekanina, & P. García-Sánchez (Eds.), Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings (pp. 213-227). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11451 LNCS). Switzerland: Springer Verlag. https://doi.org/10.1007/978-3-030-16670-0_14
Azzali, Irene ; Vanneschi, Leonardo ; Silva, Sara ; Bakurov, Illya ; Giacobini, Mario. / A Vectorial Approach to Genetic Programming. Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings. editor / Nuno Lourenço ; Ting Hu ; Hendrik Richter ; Lukas Sekanina ; Pablo García-Sánchez. Switzerland : Springer Verlag, 2019. pp. 213-227 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{21d158b800f846a68492f346af50c343,
title = "A Vectorial Approach to Genetic Programming",
abstract = "Among the various typologies of problems to which Genetic Programming (GP) has been applied since its origins, symbolic regression is one of the most popular. A common situation consists in the prediction of a target time series based on scalar features and other time series variables collected from multiple subjects. To manage this problem with GP data needs a panel representation where each observation corresponds to a collection on a subject at a precise time instant. However, representing data in this form may imply a loss of information: for instance, the algorithm may not be able to recognize observations belonging to the same subject and their recording order. To maintain the source of knowledge supplied by ordered sequences as time series, we propose a new approach to GP that keeps instances of the same observation together in a vector, introducing vectorial variables as terminals. This new representation allows aggregate functions in the primitive GP set, included with the purpose of describing the behaviour of vectorial variables. In this work, we perform a comparative analysis of vectorial GP (VE-GP) against standard GP (ST-GP). Experiments are conducted on different benchmark problems to highlight the advantages of this new approach.",
keywords = "Genetic programming, Panel data regression, Vector-based representation",
author = "Irene Azzali and Leonardo Vanneschi and Sara Silva and Illya Bakurov and Mario Giacobini",
note = "Azzali, I., Vanneschi, L., Silva, S., Bakurov, I., & Giacobini, M. (2019). A Vectorial Approach to Genetic Programming. In N. Louren{\cc}o, T. Hu, H. Richter, L. Sekanina, & P. Garc{\'i}a-S{\'a}nchez (Eds.), Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings (pp. 213-227). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11451 LNCS). Switzerland: Springer Verlag. https://doi.org/10.1007/978-3-030-16670-0_14",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-16670-0_14",
language = "English",
isbn = "9783030166694",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "213--227",
editor = "Nuno Louren{\cc}o and Ting Hu and Hendrik Richter and Lukas Sekanina and Pablo Garc{\'i}a-S{\'a}nchez",
booktitle = "Genetic Programming",

}

Azzali, I, Vanneschi, L, Silva, S, Bakurov, I & Giacobini, M 2019, A Vectorial Approach to Genetic Programming. in N Lourenço, T Hu, H Richter, L Sekanina & P García-Sánchez (eds), Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11451 LNCS, Springer Verlag, Switzerland, pp. 213-227, 22nd European Conference on Genetic Programming, EuroGP 2019, held as part of EvoStar 2019, Leipzig, Germany, 24/04/19. https://doi.org/10.1007/978-3-030-16670-0_14

A Vectorial Approach to Genetic Programming. / Azzali, Irene; Vanneschi, Leonardo; Silva, Sara; Bakurov, Illya; Giacobini, Mario.

Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings. ed. / Nuno Lourenço; Ting Hu; Hendrik Richter; Lukas Sekanina; Pablo García-Sánchez. Switzerland : Springer Verlag, 2019. p. 213-227 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11451 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A Vectorial Approach to Genetic Programming

AU - Azzali, Irene

AU - Vanneschi, Leonardo

AU - Silva, Sara

AU - Bakurov, Illya

AU - Giacobini, Mario

N1 - Azzali, I., Vanneschi, L., Silva, S., Bakurov, I., & Giacobini, M. (2019). A Vectorial Approach to Genetic Programming. In N. Lourenço, T. Hu, H. Richter, L. Sekanina, & P. García-Sánchez (Eds.), Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings (pp. 213-227). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11451 LNCS). Switzerland: Springer Verlag. https://doi.org/10.1007/978-3-030-16670-0_14

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Among the various typologies of problems to which Genetic Programming (GP) has been applied since its origins, symbolic regression is one of the most popular. A common situation consists in the prediction of a target time series based on scalar features and other time series variables collected from multiple subjects. To manage this problem with GP data needs a panel representation where each observation corresponds to a collection on a subject at a precise time instant. However, representing data in this form may imply a loss of information: for instance, the algorithm may not be able to recognize observations belonging to the same subject and their recording order. To maintain the source of knowledge supplied by ordered sequences as time series, we propose a new approach to GP that keeps instances of the same observation together in a vector, introducing vectorial variables as terminals. This new representation allows aggregate functions in the primitive GP set, included with the purpose of describing the behaviour of vectorial variables. In this work, we perform a comparative analysis of vectorial GP (VE-GP) against standard GP (ST-GP). Experiments are conducted on different benchmark problems to highlight the advantages of this new approach.

AB - Among the various typologies of problems to which Genetic Programming (GP) has been applied since its origins, symbolic regression is one of the most popular. A common situation consists in the prediction of a target time series based on scalar features and other time series variables collected from multiple subjects. To manage this problem with GP data needs a panel representation where each observation corresponds to a collection on a subject at a precise time instant. However, representing data in this form may imply a loss of information: for instance, the algorithm may not be able to recognize observations belonging to the same subject and their recording order. To maintain the source of knowledge supplied by ordered sequences as time series, we propose a new approach to GP that keeps instances of the same observation together in a vector, introducing vectorial variables as terminals. This new representation allows aggregate functions in the primitive GP set, included with the purpose of describing the behaviour of vectorial variables. In this work, we perform a comparative analysis of vectorial GP (VE-GP) against standard GP (ST-GP). Experiments are conducted on different benchmark problems to highlight the advantages of this new approach.

KW - Genetic programming

KW - Panel data regression

KW - Vector-based representation

UR - http://www.scopus.com/inward/record.url?scp=85064935476&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-16670-0_14

DO - 10.1007/978-3-030-16670-0_14

M3 - Conference contribution

SN - 9783030166694

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 213

EP - 227

BT - Genetic Programming

A2 - Lourenço, Nuno

A2 - Hu, Ting

A2 - Richter, Hendrik

A2 - Sekanina, Lukas

A2 - García-Sánchez, Pablo

PB - Springer Verlag

CY - Switzerland

ER -

Azzali I, Vanneschi L, Silva S, Bakurov I, Giacobini M. A Vectorial Approach to Genetic Programming. In Lourenço N, Hu T, Richter H, Sekanina L, García-Sánchez P, editors, Genetic Programming: 22nd European Conference, EuroGP 2019, Held as Part of EvoStar 2019, Proceedings. Switzerland: Springer Verlag. 2019. p. 213-227. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-16670-0_14