Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance

This study introduces a novel measure for evaluating attribute relevance, specifically designed to accurately identify attributes that are intrinsically related to a phenomenon, while being sensitive to the asymmetry of those relationships and noise conditions. Traditional variable selection techniq...

Full description

Saved in:
Bibliographic Details
Main Authors: Pablo Neirz, Hector Allende, Carolina Saavedra
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Algorithms
Subjects:
Online Access:https://www.mdpi.com/1999-4893/17/11/518
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850149929284009984
author Pablo Neirz
Hector Allende
Carolina Saavedra
author_facet Pablo Neirz
Hector Allende
Carolina Saavedra
author_sort Pablo Neirz
collection DOAJ
description This study introduces a novel measure for evaluating attribute relevance, specifically designed to accurately identify attributes that are intrinsically related to a phenomenon, while being sensitive to the asymmetry of those relationships and noise conditions. Traditional variable selection techniques, such as filter and wrapper methods, often fall short in capturing these complexities. Our methodology, grounded in decision trees but extendable to other machine learning models, was rigorously evaluated across various data scenarios. The results demonstrate that our measure effectively distinguishes relevant from irrelevant attributes and highlights how relevance is influenced by noise, providing a more nuanced understanding compared to established methods such as Pearson, Spearman, Kendall, MIC, MAS, MEV, GMIC, and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>P</mi><mi>h</mi><msub><mi>i</mi><mi>k</mi></msub></mrow></semantics></math></inline-formula>. This research underscores the importance of phenomenon-centric explainability, reproducibility, and robust attribute relevance evaluation in the development of predictive models. By enhancing both the interpretability and contextual accuracy of models, our approach not only supports more informed decision making but also contributes to a deeper understanding of the underlying mechanisms in diverse application domains, such as biomedical research, financial modeling, astronomy, and others.
format Article
id doaj-art-2662aec2941d477cac3e5eee8af0744a
institution OA Journals
issn 1999-4893
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Algorithms
spelling doaj-art-2662aec2941d477cac3e5eee8af0744a2025-08-20T02:26:44ZengMDPI AGAlgorithms1999-48932024-11-01171151810.3390/a17110518Attribute Relevance Score: A Novel Measure for Identifying Attribute ImportancePablo Neirz0Hector Allende1Carolina Saavedra2Departamento de Informática, Universidad Técnica Federico Santa María, Valparaíso 1680, ChileDepartamento de Informática, Universidad Técnica Federico Santa María, Valparaíso 1680, ChileDepartamento de Informática, Universidad Técnica Federico Santa María, Valparaíso 1680, ChileThis study introduces a novel measure for evaluating attribute relevance, specifically designed to accurately identify attributes that are intrinsically related to a phenomenon, while being sensitive to the asymmetry of those relationships and noise conditions. Traditional variable selection techniques, such as filter and wrapper methods, often fall short in capturing these complexities. Our methodology, grounded in decision trees but extendable to other machine learning models, was rigorously evaluated across various data scenarios. The results demonstrate that our measure effectively distinguishes relevant from irrelevant attributes and highlights how relevance is influenced by noise, providing a more nuanced understanding compared to established methods such as Pearson, Spearman, Kendall, MIC, MAS, MEV, GMIC, and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>P</mi><mi>h</mi><msub><mi>i</mi><mi>k</mi></msub></mrow></semantics></math></inline-formula>. This research underscores the importance of phenomenon-centric explainability, reproducibility, and robust attribute relevance evaluation in the development of predictive models. By enhancing both the interpretability and contextual accuracy of models, our approach not only supports more informed decision making but also contributes to a deeper understanding of the underlying mechanisms in diverse application domains, such as biomedical research, financial modeling, astronomy, and others.https://www.mdpi.com/1999-4893/17/11/518feature importancefeature selectionfeature rankingdependency measuresmachine learning explainabilityall-relevant problem
spellingShingle Pablo Neirz
Hector Allende
Carolina Saavedra
Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
Algorithms
feature importance
feature selection
feature ranking
dependency measures
machine learning explainability
all-relevant problem
title Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
title_full Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
title_fullStr Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
title_full_unstemmed Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
title_short Attribute Relevance Score: A Novel Measure for Identifying Attribute Importance
title_sort attribute relevance score a novel measure for identifying attribute importance
topic feature importance
feature selection
feature ranking
dependency measures
machine learning explainability
all-relevant problem
url https://www.mdpi.com/1999-4893/17/11/518
work_keys_str_mv AT pabloneirz attributerelevancescoreanovelmeasureforidentifyingattributeimportance
AT hectorallende attributerelevancescoreanovelmeasureforidentifyingattributeimportance
AT carolinasaavedra attributerelevancescoreanovelmeasureforidentifyingattributeimportance