{"title":"基于云的机器学习技术的水泥强度预测","authors":"Nand Kumar, V. Naranje, S. Salunkhe","doi":"10.1080/24705314.2020.1783122","DOIUrl":null,"url":null,"abstract":"ABSTRACT This paper describes a cloud-based software framework to predict cement strength for 2 days, 7 days and 28 days. Levenbarg-Marquardt back-propagation-artificial neural network (LMBP-ANN) is used to build a prediction model. This ANN model uses 70% of data for training (70%, 212 data records), testing (15%, 46 data records) and for validation (15%, 46 data records). A total of 16 significant input parameters are considered for the cement strength prediction. The user interface and software framework are built using the Python programming language. Multiple Python packages are used for the implementation of the ANN model. The cloud server having Ubuntu operating system has been used to host the web application for prediction of cement strength. The software application is tested using real-time data from various cement industries. The prediction of the cement strength of the proposed ANN-based software application appears to be very similar to those currently generated in experimental data in the cement manufacturing industry. The adequacy of the developed model based on the back-propagation ANN algorithm is confirmed as the Pearson correlation of experimental value and predicted value. The calculated value of R for experimentations on the data is 0.82539 and is 0.6813.","PeriodicalId":43844,"journal":{"name":"Journal of Structural Integrity and Maintenance","volume":null,"pages":null},"PeriodicalIF":3.0000,"publicationDate":"2020-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24705314.2020.1783122","citationCount":"9","resultStr":"{\"title\":\"Cement strength prediction using cloud-based machine learning techniques\",\"authors\":\"Nand Kumar, V. Naranje, S. Salunkhe\",\"doi\":\"10.1080/24705314.2020.1783122\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"ABSTRACT This paper describes a cloud-based software framework to predict cement strength for 2 days, 7 days and 28 days. Levenbarg-Marquardt back-propagation-artificial neural network (LMBP-ANN) is used to build a prediction model. This ANN model uses 70% of data for training (70%, 212 data records), testing (15%, 46 data records) and for validation (15%, 46 data records). A total of 16 significant input parameters are considered for the cement strength prediction. The user interface and software framework are built using the Python programming language. Multiple Python packages are used for the implementation of the ANN model. The cloud server having Ubuntu operating system has been used to host the web application for prediction of cement strength. The software application is tested using real-time data from various cement industries. The prediction of the cement strength of the proposed ANN-based software application appears to be very similar to those currently generated in experimental data in the cement manufacturing industry. The adequacy of the developed model based on the back-propagation ANN algorithm is confirmed as the Pearson correlation of experimental value and predicted value. The calculated value of R for experimentations on the data is 0.82539 and is 0.6813.\",\"PeriodicalId\":43844,\"journal\":{\"name\":\"Journal of Structural Integrity and Maintenance\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.0000,\"publicationDate\":\"2020-09-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://sci-hub-pdf.com/10.1080/24705314.2020.1783122\",\"citationCount\":\"9\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Structural Integrity and Maintenance\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1080/24705314.2020.1783122\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Structural Integrity and Maintenance","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1080/24705314.2020.1783122","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Cement strength prediction using cloud-based machine learning techniques
ABSTRACT This paper describes a cloud-based software framework to predict cement strength for 2 days, 7 days and 28 days. Levenbarg-Marquardt back-propagation-artificial neural network (LMBP-ANN) is used to build a prediction model. This ANN model uses 70% of data for training (70%, 212 data records), testing (15%, 46 data records) and for validation (15%, 46 data records). A total of 16 significant input parameters are considered for the cement strength prediction. The user interface and software framework are built using the Python programming language. Multiple Python packages are used for the implementation of the ANN model. The cloud server having Ubuntu operating system has been used to host the web application for prediction of cement strength. The software application is tested using real-time data from various cement industries. The prediction of the cement strength of the proposed ANN-based software application appears to be very similar to those currently generated in experimental data in the cement manufacturing industry. The adequacy of the developed model based on the back-propagation ANN algorithm is confirmed as the Pearson correlation of experimental value and predicted value. The calculated value of R for experimentations on the data is 0.82539 and is 0.6813.