Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems

Over the recent years, Industry 4.0 (I4.0) technologies such as the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and the presence of Industrial Big Data (IBD) have helped achieve intelligent Fault Detection (FD) in manufacturing. Notably, data-driven approaches in FD apply Dee...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Jefkine Kafunah, Muhammad Intizar Ali, John G. Breslin
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	fault detection imbalanced datasets deep neural networks Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/df2717f2ba0d4f4b9e166c8410039dc0
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:df2717f2ba0d4f4b9e166c8410039dc0
record_format	dspace
spelling	oai:doaj.org-article:df2717f2ba0d4f4b9e166c8410039dc02021-11-11T14:56:54ZHandling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems10.3390/app112197832076-3417https://doaj.org/article/df2717f2ba0d4f4b9e166c8410039dc02021-10-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/21/9783https://doaj.org/toc/2076-3417Over the recent years, Industry 4.0 (I4.0) technologies such as the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and the presence of Industrial Big Data (IBD) have helped achieve intelligent Fault Detection (FD) in manufacturing. Notably, data-driven approaches in FD apply Deep Learning (DL) techniques to help generate insights required for monitoring complex manufacturing processes. However, due to the ratio of instances where actual faults occur, FD datasets tend to be imbalanced, leading to training challenges that result in inefficient DL-based FD models. In this paper, we propose Dual Logits Weights Perturbation (DLWP) loss, a method featuring weight vectors for improved dataset generalization in FD systems. The weight vectors act as hyperparameters adjusted on a case-by-case basis to regulate focus accorded to individual minority classes during training. In particular, our proposed method is suitable for imbalanced datasets from safety-related FD tasks as it generates DL models that minimize false negatives. Subsequently, we integrate human experts into the workflow as a strategy to help safeguard the system. A subset of the results, model predictions with uncertainties exceeding a preset threshold, are considered a preliminary output subject to cross-checking by human experts. We demonstrate that DLWP achieves improved Recall, AUC, F1 scores.Jefkine KafunahMuhammad Intizar AliJohn G. BreslinMDPI AGarticlefault detectionimbalanced datasetsdeep neural networksTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 9783, p 9783 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	fault detection imbalanced datasets deep neural networks Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	fault detection imbalanced datasets deep neural networks Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Jefkine Kafunah Muhammad Intizar Ali John G. Breslin Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
description	Over the recent years, Industry 4.0 (I4.0) technologies such as the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and the presence of Industrial Big Data (IBD) have helped achieve intelligent Fault Detection (FD) in manufacturing. Notably, data-driven approaches in FD apply Deep Learning (DL) techniques to help generate insights required for monitoring complex manufacturing processes. However, due to the ratio of instances where actual faults occur, FD datasets tend to be imbalanced, leading to training challenges that result in inefficient DL-based FD models. In this paper, we propose Dual Logits Weights Perturbation (DLWP) loss, a method featuring weight vectors for improved dataset generalization in FD systems. The weight vectors act as hyperparameters adjusted on a case-by-case basis to regulate focus accorded to individual minority classes during training. In particular, our proposed method is suitable for imbalanced datasets from safety-related FD tasks as it generates DL models that minimize false negatives. Subsequently, we integrate human experts into the workflow as a strategy to help safeguard the system. A subset of the results, model predictions with uncertainties exceeding a preset threshold, are considered a preliminary output subject to cross-checking by human experts. We demonstrate that DLWP achieves improved Recall, AUC, F1 scores.
format	article
author	Jefkine Kafunah Muhammad Intizar Ali John G. Breslin
author_facet	Jefkine Kafunah Muhammad Intizar Ali John G. Breslin
author_sort	Jefkine Kafunah
title	Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
title_short	Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
title_full	Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
title_fullStr	Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
title_full_unstemmed	Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems
title_sort	handling imbalanced datasets for robust deep neural network-based fault detection in manufacturing systems
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/df2717f2ba0d4f4b9e166c8410039dc0
work_keys_str_mv	AT jefkinekafunah handlingimbalanceddatasetsforrobustdeepneuralnetworkbasedfaultdetectioninmanufacturingsystems AT muhammadintizarali handlingimbalanceddatasetsforrobustdeepneuralnetworkbasedfaultdetectioninmanufacturingsystems AT johngbreslin handlingimbalanceddatasetsforrobustdeepneuralnetworkbasedfaultdetectioninmanufacturingsystems
_version_	1718437935975497728

Handling Imbalanced Datasets for Robust Deep Neural Network-Based Fault Detection in Manufacturing Systems

Ejemplares similares