Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement

ABSTRACT: Non-metallic inclusions are unavoidably produced during steel casting resulting in lower mechanical strength and other detrimental effects. This study was aimed at developing a reliable Machine Learning algorithm to classify castings of steel for tire reinforcement depending on the number...

Descripción completa

Detalles Bibliográficos
Autor: Ruiz Martínez, Estela
Tipo de recurso: tesis de maestría
Fecha de publicación:2019
País:España
Institución:Universidad de Cantabria (UC)
Repositorio:UCrea Repositorio Abierto de la Universidad de Cantabria
Idioma:inglés
OAI Identifier:oai:repositorio.unican.es:10902/16903
Acceso en línea:http://hdl.handle.net/10902/16903
Access Level:acceso abierto
Palabra clave:Machine learning
Steel wire
Continuous casting
Non-metallic inclusions
Random Forest
Imbalanced dataset
Aprendizaje automático
Alambrón de acero
Colada continua
Inclusiones no metálicas
Conjunto de datos desequilibrado
id ES_8c82c55c0eef7d4ff8b5fbe4fa88eefe
oai_identifier_str oai:repositorio.unican.es:10902/16903
network_acronym_str ES
network_name_str España
repository_id_str
dc.title.none.fl_str_mv Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
Métodos machine learning para la predicción de inclusiones no metálicas en alambres de acero para refuerzo de neumáticos
title Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
spellingShingle Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
Ruiz Martínez, Estela
Machine learning
Steel wire
Continuous casting
Non-metallic inclusions
Random Forest
Imbalanced dataset
Aprendizaje automático
Alambrón de acero
Colada continua
Inclusiones no metálicas
Conjunto de datos desequilibrado
title_short Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
title_full Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
title_fullStr Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
title_full_unstemmed Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
title_sort Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcement
dc.creator.none.fl_str_mv Ruiz Martínez, Estela
author Ruiz Martínez, Estela
author_facet Ruiz Martínez, Estela
author_role author
dc.contributor.none.fl_str_mv Lloret Iglesias, Lara
Universidad de Cantabria
dc.subject.none.fl_str_mv Machine learning
Steel wire
Continuous casting
Non-metallic inclusions
Random Forest
Imbalanced dataset
Aprendizaje automático
Alambrón de acero
Colada continua
Inclusiones no metálicas
Conjunto de datos desequilibrado
topic Machine learning
Steel wire
Continuous casting
Non-metallic inclusions
Random Forest
Imbalanced dataset
Aprendizaje automático
Alambrón de acero
Colada continua
Inclusiones no metálicas
Conjunto de datos desequilibrado
description ABSTRACT: Non-metallic inclusions are unavoidably produced during steel casting resulting in lower mechanical strength and other detrimental effects. This study was aimed at developing a reliable Machine Learning algorithm to classify castings of steel for tire reinforcement depending on the number and properties of inclusions, experimentally determined. 855 observations were available for training, validation and testing the algorithms, obtained from the quality control of the steel. 140 parameters are monitored during fabrication, which are the features of the analysis; the output is 1 or 0 depending on whether the casting is rejected or not. The following algorithms have been employed: Logistic Regression, K-Nearest Neighbors, Support Vector Classifier (linear and RBF kernels), Random Forests, AdaBoost, Gradient Boosting and Artificial Neural Networks. The reduced value of the rejection rate implies that classification must be carried out on an imbalanced dataset. Resampling methods and specific scores for imbalanced datasets (Recall, Precision and AUC rather than Accuracy) were used. Random Forest was the most successful method providing an AUC in the test set of 0.85. No significant improvements were detected after resampling. The improvement derived from implementing this algorithm in the sampling procedure for quality control during steelmaking has been quantified. In this sense, it has been proved that this tool allows the samples with a higher probability of being rejected to be selected, thus improving the effectiveness of the quality control. In addition, the optimized Random Forest has enabled to identify the most important features, which have been satisfactorily interpreted on a metallurgical basis.
publishDate 2019
dc.date.none.fl_str_mv 2019
2019-07-01
dc.type.none.fl_str_mv master thesis
http://purl.org/coar/resource_type/c_bdcc
NA
http://purl.org/coar/version/c_be7fb7dd8ff6fe43
dc.type.openaire.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
dc.identifier.none.fl_str_mv http://hdl.handle.net/10902/16903
url http://hdl.handle.net/10902/16903
dc.language.none.fl_str_mv Inglés
eng
language_invalid_str_mv Inglés
language eng
dc.rights.none.fl_str_mv open access
http://purl.org/coar/access_right/c_abf2
Atribución-NoComercial-SinDerivadas 3.0 España
http://creativecommons.org/licenses/by-nc-nd/3.0/es/
dc.rights.openaire.fl_str_mv info:eu-repo/semantics/openAccess
rights_invalid_str_mv open access
http://purl.org/coar/access_right/c_abf2
Atribución-NoComercial-SinDerivadas 3.0 España
http://creativecommons.org/licenses/by-nc-nd/3.0/es/
eu_rights_str_mv openAccess
dc.source.none.fl_str_mv reponame:UCrea Repositorio Abierto de la Universidad de Cantabria
instname:Universidad de Cantabria (UC)
instname_str Universidad de Cantabria (UC)
reponame_str UCrea Repositorio Abierto de la Universidad de Cantabria
collection UCrea Repositorio Abierto de la Universidad de Cantabria
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869412938958241793
spelling Machine learning methods for the prediction of non-metallic inclusions in steel wires for tire reinforcementMétodos machine learning para la predicción de inclusiones no metálicas en alambres de acero para refuerzo de neumáticosRuiz Martínez, EstelaMachine learningSteel wireContinuous castingNon-metallic inclusionsRandom ForestImbalanced datasetAprendizaje automáticoAlambrón de aceroColada continuaInclusiones no metálicasConjunto de datos desequilibradoABSTRACT: Non-metallic inclusions are unavoidably produced during steel casting resulting in lower mechanical strength and other detrimental effects. This study was aimed at developing a reliable Machine Learning algorithm to classify castings of steel for tire reinforcement depending on the number and properties of inclusions, experimentally determined. 855 observations were available for training, validation and testing the algorithms, obtained from the quality control of the steel. 140 parameters are monitored during fabrication, which are the features of the analysis; the output is 1 or 0 depending on whether the casting is rejected or not. The following algorithms have been employed: Logistic Regression, K-Nearest Neighbors, Support Vector Classifier (linear and RBF kernels), Random Forests, AdaBoost, Gradient Boosting and Artificial Neural Networks. The reduced value of the rejection rate implies that classification must be carried out on an imbalanced dataset. Resampling methods and specific scores for imbalanced datasets (Recall, Precision and AUC rather than Accuracy) were used. Random Forest was the most successful method providing an AUC in the test set of 0.85. No significant improvements were detected after resampling. The improvement derived from implementing this algorithm in the sampling procedure for quality control during steelmaking has been quantified. In this sense, it has been proved that this tool allows the samples with a higher probability of being rejected to be selected, thus improving the effectiveness of the quality control. In addition, the optimized Random Forest has enabled to identify the most important features, which have been satisfactorily interpreted on a metallurgical basis.RESUMEN: Las inclusiones no metálicas se producen inevitablemente durante la fabricación del acero, lo que resulta en una menor resistencia mecánica y otros efectos perjudiciales. El objetivo de este estudio fue desarrollar un algoritmo fiable para clasificar las coladas de acero de refuerzo de neumáticos en función del número y el tipo de las inclusiones, determinadas experimentalmente. Se dispuso de 855 observaciones para el entrenamiento, validación y test de los algoritmos, obtenidos a partir del control de calidad del acero. Durante la fabricación se controlan 140 parámetros, que son las características del análisis; el resultado es 1 ó 0 dependiendo de si la colada es rechazada o no. Se han empleado los siguientes algoritmos: Regresión Logística, Vecinos K-Cercanos, Clasificador de Vectores Soporte (kernels lineales y RBF), Bosques Aleatorios, AdaBoost, Gradient Boosting y Redes Neurales Artificiales. El bajo índice de rechazo implica que la clasificación debe llevarse a cabo en un set de datos desequilibrado. Se utilizaron métodos de remuestreo y métricas específicas para conjuntos de datos desequilibrados (Recall, Precision y AUC en lugar de Accuracy). Random Forest fue el algoritmo más exitoso que proporcionó un AUC en los datos de test de 0.83. No se detectaron mejoras significativas después del remuestreo. Se ha cuantificado la mejora derivada de la implementación de este algoritmo en el procedimiento de muestreo para el control de calidad durante la fabricación de acero. En este sentido, se ha comprobado que esta herramienta permite seleccionar las muestras con mayor probabilidad de ser rechazadas, mejorando así la eficacia del control de calidad. Además, el Random Forest optimizado ha permitido identificar las variables más importantes, que han sido interpretadas satisfactoriamente sobre una base metalúrgica.Máster en Ciencia de DatosLloret Iglesias, LaraUniversidad de Cantabria20192019-07-01master thesishttp://purl.org/coar/resource_type/c_bdccNAhttp://purl.org/coar/version/c_be7fb7dd8ff6fe43info:eu-repo/semantics/masterThesishttp://hdl.handle.net/10902/16903reponame:UCrea Repositorio Abierto de la Universidad de Cantabriainstname:Universidad de Cantabria (UC)Inglésengopen accesshttp://purl.org/coar/access_right/c_abf2Atribución-NoComercial-SinDerivadas 3.0 Españahttp://creativecommons.org/licenses/by-nc-nd/3.0/es/info:eu-repo/semantics/openAccessoai:repositorio.unican.es:10902/169032026-06-02T12:39:31Z
score 15,300719

Similares en LA Referencia