Show simple item record

dc.contributor.authorCarrasco Castillo, Jacinto 
dc.contributor.authorLópez Pretel, David 
dc.contributor.authorAguilera Martos, Ignacio 
dc.contributor.authorGarcía Gil, Diego Jesús 
dc.contributor.authorLuengo Martín, Julián 
dc.contributor.authorHerrera Triguero, Francisco 
dc.date.accessioned2021-10-22T06:29:23Z
dc.date.available2021-10-22T06:29:23Z
dc.date.issued2021-09-02
dc.identifier.citationPublished version: Jacinto Carrasco... [et al.]. Anomaly detection in predictive maintenance: A new evaluation framework for temporal unsupervised anomaly detection algorithms, Neurocomputing, Volume 462, 2021, Pages 440-452, ISSN 0925-2312, [https://doi.org/10.1016/j.neucom.2021.07.095]es_ES
dc.identifier.urihttp://hdl.handle.net/10481/71033
dc.descriptionThis work has been partially supported by the Ministry of Science and Technology under project TIN2017-89517-P, the Contract UGR-AM OTRI-4260 and the Andalusian Excellence project P18-FR-4961. J. Carrasco was supported by the Spanish Ministry of Science under the FPU Programme 998758-2016. D. Garcia-Gil holds a contract co-financed by the European Social Fund and the Administration of the Junta de Andalucia, reference DOC_01137.es_ES
dc.description.abstractThe research in anomaly detection lacks a unified definition of what represents an anomalous instance. Discrepancies in the nature itself of an anomaly lead to multiple paradigms of algorithms design and experimentation. Predictive maintenance is a special case, where the anomaly represents a failure that must be prevented. Related time series research as outlier and novelty detection or time series classification does not apply to the concept of an anomaly in this field, because they are not single points which have not been seen previously and may not be precisely annotated. Moreover, due to the lack of annotated anomalous data, many benchmarks are adapted from supervised scenarios. To address these issues, we generalise the concept of positive and negative instances to intervals to be able to evaluate unsupervised anomaly detection algorithms. We also preserve the imbalance scheme for evaluation through the proposal of the Preceding Window ROC, a generalisation for the calculation of ROC curves for time series scenarios. We also adapt the mechanism from a established time series anomaly detection benchmark to the proposed generalisations to reward early detection. Therefore, the proposal represents a flexible evaluation framework for the different scenarios. To show the usefulness of this definition, we include a case study of Big Data algorithms with a real-world time series problem provided by the company ArcelorMittal, and compare the proposal with an evaluation method.es_ES
dc.description.sponsorshipMinistry of Science and Technology TIN2017-89517-Pes_ES
dc.description.sponsorshipContract UGR-AM OTRI-4260es_ES
dc.description.sponsorshipAndalusian Excellence project P18-FR-4961es_ES
dc.description.sponsorshipSpanish Government 998758-2016es_ES
dc.description.sponsorshipEuropean Social Fund (ESF)es_ES
dc.description.sponsorshipJunta de Andalucia DOC_01137es_ES
dc.language.isoenges_ES
dc.publisherElsevieres_ES
dc.rightsAtribución-NoComercial-SinDerivadas 3.0 España*
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/3.0/es/*
dc.subjectAnomalyes_ES
dc.subjectOutlieres_ES
dc.subjectScore systemes_ES
dc.subjectEvaluation es_ES
dc.subjectBenchmarkes_ES
dc.titleAnomaly Detection in Predictive Maintenance: A New Evaluation Framework for Temporal Unsupervised Anomaly Detection Algorithmses_ES
dc.typeinfo:eu-repo/semantics/articlees_ES
dc.rights.accessRightsinfo:eu-repo/semantics/openAccesses_ES
dc.type.hasVersioninfo:eu-repo/semantics/submittedVersiones_ES


Files in this item

[PDF]

This item appears in the following Collection(s)

Show simple item record

Atribución-NoComercial-SinDerivadas 3.0 España
Except where otherwise noted, this item's license is described as Atribución-NoComercial-SinDerivadas 3.0 España