An Application of Machine Learning Algorithms on the Prediction of the Damage Level of Rubble-Mound Breakwaters
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Standard
An Application of Machine Learning Algorithms on the Prediction of the Damage Level of Rubble-Mound Breakwaters. / Saha, Susmita; De, Soumen; Changdar, Satyasaran.
I: Journal of Offshore Mechanics and Arctic Engineering, Bind 146, Nr. 1, 011202, 2024.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - An Application of Machine Learning Algorithms on the Prediction of the Damage Level of Rubble-Mound Breakwaters
AU - Saha, Susmita
AU - De, Soumen
AU - Changdar, Satyasaran
N1 - Publisher Copyright: Copyright © 2023 by ASME.
PY - 2024
Y1 - 2024
N2 - The stability analysis of breakwaters is very important to have a safe and economic design of these coastal protective structures and the damage level is one of the most important parameters in this context. In the recent past, machine learning techniques showed immense potential in transforming many industries and processes, for making them more efficient and accurate. In this study, five advanced machine learning algorithms, support vector regression, random forest, Adaboost, gradient boosting, and deep artificial neural network, were employed and analyzed on estimation of the damage level of rubble-mound breakwaters. A large experimental dataset, considering almost every stability variable with their whole ranges, was used in this purpose. Also, a detailed feature analysis is presented to have an insight into the relations between these variables. It was found that the present study had overcome all of the limitations of existing studies related to this field and delivered the highest level of accuracy.
AB - The stability analysis of breakwaters is very important to have a safe and economic design of these coastal protective structures and the damage level is one of the most important parameters in this context. In the recent past, machine learning techniques showed immense potential in transforming many industries and processes, for making them more efficient and accurate. In this study, five advanced machine learning algorithms, support vector regression, random forest, Adaboost, gradient boosting, and deep artificial neural network, were employed and analyzed on estimation of the damage level of rubble-mound breakwaters. A large experimental dataset, considering almost every stability variable with their whole ranges, was used in this purpose. Also, a detailed feature analysis is presented to have an insight into the relations between these variables. It was found that the present study had overcome all of the limitations of existing studies related to this field and delivered the highest level of accuracy.
KW - breakwater’s stability
KW - damage level
KW - ensemble learning
KW - feature analysis
KW - machine learning
KW - neural network
KW - regression analysis
UR - http://www.scopus.com/inward/record.url?scp=85176793325&partnerID=8YFLogxK
U2 - 10.1115/1.4062475
DO - 10.1115/1.4062475
M3 - Journal article
AN - SCOPUS:85176793325
VL - 146
JO - Journal of Offshore Mechanics and Arctic Engineering
JF - Journal of Offshore Mechanics and Arctic Engineering
SN - 0892-7219
IS - 1
M1 - 011202
ER -
ID: 374120894