Online Numerical Association Rule Miner

Fuente: Zenodo
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Galvez Tomida, Akemi, Iglesias, Andres
Format: Recurso digital
Veröffentlicht: Zenodo 2022
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866901948910796800
author Galvez Tomida, Akemi
Iglesias, Andres
author_facet Galvez Tomida, Akemi
Iglesias, Andres
contents <p>Green AI refers to those AI methods that are friendly to the environment, i.e., are capable to keep the consumption of electrical energy at a minimum. In this sense, a new numerical association rule miner is proposed that presents a combination of the already existing offline uARMSolver, belonging to a Red AI class, and a newly developed onlineNARM miner representing the new Green AI. The former is devoted to exhaustive search of the evolutionary solution space, while the latter for faster exploiting of already explored search space. The experimental results on four transaction databases showed that, by sacrificing the quality of the results by 0.7 %, by the onlineNARM we can obtain the results almost 85.0 % faster than with the uARMSolver in the best test scenario</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_14263009
institution Zenodo
language
publishDate 2022
publisher Zenodo
record_format zenodo
spellingShingle Online Numerical Association Rule Miner
Galvez Tomida, Akemi
Iglesias, Andres
Green AI, Red AI, numerical association rule mining, uARMSolver, onlineNARM.
<p>Green AI refers to those AI methods that are friendly to the environment, i.e., are capable to keep the consumption of electrical energy at a minimum. In this sense, a new numerical association rule miner is proposed that presents a combination of the already existing offline uARMSolver, belonging to a Red AI class, and a newly developed onlineNARM miner representing the new Green AI. The former is devoted to exhaustive search of the evolutionary solution space, while the latter for faster exploiting of already explored search space. The experimental results on four transaction databases showed that, by sacrificing the quality of the results by 0.7 %, by the onlineNARM we can obtain the results almost 85.0 % faster than with the uARMSolver in the best test scenario</p>
title Online Numerical Association Rule Miner
topic Green AI, Red AI, numerical association rule mining, uARMSolver, onlineNARM.
url https://doi.org/10.5281/zenodo.14263009