A regression-like classification system for geometric semantic genetic programming

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

Abstract

Geometric Semantic Genetic Programming (GSGP) is a recent variant of Genetic Programming, that is gaining popularity thanks to its ability to induce a unimodal error surface for any supervised learning problem. Nevertheless, so far GSGP has been applied to the real world basically only on regression problems. This paper represents an attempt to apply GSGP to real world classification problems. Taking inspiration from Per-ceptron neural networks, we represent class labels as numbers and we use an activation function to constraint the output of the solutions in a given range of possible values. In this way, the classification problem is turned into a regression one, and traditional GSGP can be used. In this work, we focus on binary classification; logistic constraining outputs in [0,1] is used as an activation function and the class labels are transformed into 0 and 1. The use of the logistic activation function helps to improve the generalization ability of the system. The presented results are encouraging: our regression-based classification system was able to obtain results that are better than, or comparable to, the ones of a set of competitor machine learning methods, on a rather rich set of real-life test problems.

Original languageEnglish
Title of host publicationProceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019)
EditorsJuan Julian Merelo, Jonathan Garibaldi, Alejandro Linares-Barranco, Kurosh Madani, Kevin Warwick, Kevin Warwick
PublisherSciTePress
Pages40-48
Number of pages9
Volume1
ISBN (Electronic)9789897583841
Publication statusPublished - 1 Jan 2019
Event11th International Joint Conference on Computational Intelligence, IJCCI 2019 - Vienna, Austria
Duration: 17 Sep 201919 Sep 2019

Publication series

NameIJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence

Conference

Conference11th International Joint Conference on Computational Intelligence, IJCCI 2019
CountryAustria
CityVienna
Period17/09/1919/09/19

Fingerprint

Genetic programming
Semantics
Chemical activation
Logistics
Labels
Supervised learning
Learning systems
Neural networks

Keywords

  • Classification
  • Geometric semantic genetic programming
  • Regression

Cite this

Bakurov, I., Castelli, M., Fontanella, F., & Vanneschi, L. (2019). A regression-like classification system for geometric semantic genetic programming. In J. J. Merelo, J. Garibaldi, A. Linares-Barranco, K. Madani, K. Warwick, & K. Warwick (Eds.), Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019) (Vol. 1, pp. 40-48). (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence). SciTePress.
Bakurov, Illya ; Castelli, Mauro ; Fontanella, Francesco ; Vanneschi, Leonardo. / A regression-like classification system for geometric semantic genetic programming. Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019). editor / Juan Julian Merelo ; Jonathan Garibaldi ; Alejandro Linares-Barranco ; Kurosh Madani ; Kevin Warwick ; Kevin Warwick. Vol. 1 SciTePress, 2019. pp. 40-48 (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence).
@inproceedings{36026badae8d446b990d8211446661aa,
title = "A regression-like classification system for geometric semantic genetic programming",
abstract = "Geometric Semantic Genetic Programming (GSGP) is a recent variant of Genetic Programming, that is gaining popularity thanks to its ability to induce a unimodal error surface for any supervised learning problem. Nevertheless, so far GSGP has been applied to the real world basically only on regression problems. This paper represents an attempt to apply GSGP to real world classification problems. Taking inspiration from Per-ceptron neural networks, we represent class labels as numbers and we use an activation function to constraint the output of the solutions in a given range of possible values. In this way, the classification problem is turned into a regression one, and traditional GSGP can be used. In this work, we focus on binary classification; logistic constraining outputs in [0,1] is used as an activation function and the class labels are transformed into 0 and 1. The use of the logistic activation function helps to improve the generalization ability of the system. The presented results are encouraging: our regression-based classification system was able to obtain results that are better than, or comparable to, the ones of a set of competitor machine learning methods, on a rather rich set of real-life test problems.",
keywords = "Classification, Geometric semantic genetic programming, Regression",
author = "Illya Bakurov and Mauro Castelli and Francesco Fontanella and Leonardo Vanneschi",
note = "Bakurov, I., Castelli, M., Fontanella, F., & Vanneschi, L. (2019). A regression-like classification system for geometric semantic genetic programming. In J. J. Merelo, J. Garibaldi, A. Linares-Barranco, K. Madani, K. Warwick, & K. Warwick (Eds.), Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019) (Vol. 1, pp. 40-48). (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence). SciTePress.",
year = "2019",
month = "1",
day = "1",
language = "English",
volume = "1",
series = "IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence",
publisher = "SciTePress",
pages = "40--48",
editor = "Merelo, {Juan Julian} and Jonathan Garibaldi and Alejandro Linares-Barranco and Kurosh Madani and Kevin Warwick and Kevin Warwick",
booktitle = "Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019)",

}

Bakurov, I, Castelli, M, Fontanella, F & Vanneschi, L 2019, A regression-like classification system for geometric semantic genetic programming. in JJ Merelo, J Garibaldi, A Linares-Barranco, K Madani, K Warwick & K Warwick (eds), Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019). vol. 1, IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence, SciTePress, pp. 40-48, 11th International Joint Conference on Computational Intelligence, IJCCI 2019, Vienna, Austria, 17/09/19.

A regression-like classification system for geometric semantic genetic programming. / Bakurov, Illya; Castelli, Mauro; Fontanella, Francesco; Vanneschi, Leonardo.

Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019). ed. / Juan Julian Merelo; Jonathan Garibaldi; Alejandro Linares-Barranco; Kurosh Madani; Kevin Warwick; Kevin Warwick. Vol. 1 SciTePress, 2019. p. 40-48 (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence).

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

TY - GEN

T1 - A regression-like classification system for geometric semantic genetic programming

AU - Bakurov, Illya

AU - Castelli, Mauro

AU - Fontanella, Francesco

AU - Vanneschi, Leonardo

N1 - Bakurov, I., Castelli, M., Fontanella, F., & Vanneschi, L. (2019). A regression-like classification system for geometric semantic genetic programming. In J. J. Merelo, J. Garibaldi, A. Linares-Barranco, K. Madani, K. Warwick, & K. Warwick (Eds.), Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019) (Vol. 1, pp. 40-48). (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence). SciTePress.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Geometric Semantic Genetic Programming (GSGP) is a recent variant of Genetic Programming, that is gaining popularity thanks to its ability to induce a unimodal error surface for any supervised learning problem. Nevertheless, so far GSGP has been applied to the real world basically only on regression problems. This paper represents an attempt to apply GSGP to real world classification problems. Taking inspiration from Per-ceptron neural networks, we represent class labels as numbers and we use an activation function to constraint the output of the solutions in a given range of possible values. In this way, the classification problem is turned into a regression one, and traditional GSGP can be used. In this work, we focus on binary classification; logistic constraining outputs in [0,1] is used as an activation function and the class labels are transformed into 0 and 1. The use of the logistic activation function helps to improve the generalization ability of the system. The presented results are encouraging: our regression-based classification system was able to obtain results that are better than, or comparable to, the ones of a set of competitor machine learning methods, on a rather rich set of real-life test problems.

AB - Geometric Semantic Genetic Programming (GSGP) is a recent variant of Genetic Programming, that is gaining popularity thanks to its ability to induce a unimodal error surface for any supervised learning problem. Nevertheless, so far GSGP has been applied to the real world basically only on regression problems. This paper represents an attempt to apply GSGP to real world classification problems. Taking inspiration from Per-ceptron neural networks, we represent class labels as numbers and we use an activation function to constraint the output of the solutions in a given range of possible values. In this way, the classification problem is turned into a regression one, and traditional GSGP can be used. In this work, we focus on binary classification; logistic constraining outputs in [0,1] is used as an activation function and the class labels are transformed into 0 and 1. The use of the logistic activation function helps to improve the generalization ability of the system. The presented results are encouraging: our regression-based classification system was able to obtain results that are better than, or comparable to, the ones of a set of competitor machine learning methods, on a rather rich set of real-life test problems.

KW - Classification

KW - Geometric semantic genetic programming

KW - Regression

UR - http://www.scopus.com/inward/record.url?scp=85074291742&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1

T3 - IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence

SP - 40

EP - 48

BT - Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019)

A2 - Merelo, Juan Julian

A2 - Garibaldi, Jonathan

A2 - Linares-Barranco, Alejandro

A2 - Madani, Kurosh

A2 - Warwick, Kevin

A2 - Warwick, Kevin

PB - SciTePress

ER -

Bakurov I, Castelli M, Fontanella F, Vanneschi L. A regression-like classification system for geometric semantic genetic programming. In Merelo JJ, Garibaldi J, Linares-Barranco A, Madani K, Warwick K, Warwick K, editors, Proceedings of the 11th International Joint Conference on Computational Intelligence (IJCCI 2019). Vol. 1. SciTePress. 2019. p. 40-48. (IJCCI 2019 - Proceedings of the 11th International Joint Conference on Computational Intelligence).