Author(s):
Vanneschi, Leonardo ; Castelli, Mauro ; Manzoni, Luca ; Silva, Sara ; Trujillo, Leonardo
Date: 2020
Persistent ID: http://hdl.handle.net/10362/146560
Origin: Repositório Institucional da UNL
Project/scholarship:
info:eu-repo/grantAgreement/FCT/9471 - RIDTI/PTDC%2FCCI-CIF%2F29877%2F2017/PT;
info:eu-repo/grantAgreement/FCT/3599-PPCDT/DSAIPA%2FDS%2F0113%2F2019/PT;
info:eu-repo/grantAgreement/FCT/3599-PPCDT/DSAIPA%2FDS%2F0022%2F2018/PT;
info:eu-repo/grantAgreement/FCT/3599-PPCDT/PTDC%2FCCI-INF%2F29168%2F2017/PT;
Subject(s): Theoretical Computer Science; Computer Science(all)
Description
Vanneschi, L., Castelli, M., Manzoni, L., Silva, S., & Trujillo, L. (2020). Is k Nearest Neighbours Regression Better Than GP? In T. Hu, N. Lourenço, E. Medvet, & F. Divina (Eds.), Genetic Programming - 23rd European Conference, EuroGP 2020, Held as Part of EvoStar 2020, Proceedings (pp. 244-261). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12101 LNCS). Springer. https://doi.org/10.1007/978-3-030-44094-7_16
This work starts from the empirical observation that k nearest neighbours (KNN) consistently outperforms state-of-the-art techniques for regression, including geometric semantic genetic programming (GSGP). However, KNN is a memorization, and not a learning, method, i.e. it evaluates unseen data on the basis of training observations, and not by running a learned model. This paper takes a first step towards the objective of defining a learning method able to equal KNN, by defining a new semantic mutation, called random vectors-based mutation (RVM). GP using RVM, called RVMGP, obtains results that are comparable to KNN, but still needs training data to evaluate unseen instances. A comparative analysis sheds some light on the reason why RVMGP outperforms GSGP, revealing that RVMGP is able to explore the semantic space more uniformly. This finding opens a question for the future: is it possible to define a new genetic operator, that explores the semantic space as uniformly as RVM does, but that still allows us to evaluate unseen instances without using training data?
