Performance Comparison of Model Storage Formats for Deploying Data Mining Models
| dc.authorid | 0000-0001-6698-8726 | en_US |
| dc.authorid | 0000-0002-9245-5728 | en_US |
| dc.contributor.author | Yıldız, Ersin | |
| dc.contributor.author | Bilgin, Turgay Tugay | |
| dc.date.accessioned | 2023-03-18T18:21:14Z | |
| dc.date.available | 2023-03-18T18:21:14Z | |
| dc.date.issued | 2022 | en_US |
| dc.department | Başka Kurum | en_US |
| dc.description.abstract | Electronics and computer technology are rapidly changing. This trend has both changed the habits of the end-users and initiated a transformation in the industry. In addition, data collection, storage, and processing studies have gained importance with Industry 4.0 (I4.0). Nowadays, data has become an indispensable resource for information extraction. Recently, predictive models produced by machine learning and data mining have been frequently used in our daily life. The wide variety of environments in which these models can be developed is easy for developers while integrating models into existing systems is equally difficult and costly. In this study, two approaches are compared in terms of their performance in publishing model files serialized with predictive model markup language, which is accepted by frequently used software tools such as Knime, Weka, IBM SPSS Modeler. Python programming language and tools are used in interpreting predictive model markup language files and creating web services. The first approach was chosen to store predictive model markup language file contents in a database, and the second approach was to store these files in the file system, and these two methods were measured and reported on the parameters of response time, throughput (speed of processing requests) and latency in generating responses. As a result of the measurements made, it has been seen that the web service performance is higher when the model is kept as a file in the file system compared to the other method. | en_US |
| dc.identifier.citation | Yıldız, E., Bilgin, T. T., (2022). Performance Comparison of Model Storage Formats for Deploying Data Mining Models. Selcuk University Journal of Engineering Sciences, 21, (02), 52-57. | en_US |
| dc.identifier.endpage | 57 | en_US |
| dc.identifier.issn | 2757-8828 | en_US |
| dc.identifier.issue | 02 | en_US |
| dc.identifier.startpage | 52 | en_US |
| dc.identifier.uri | https://hdl.handle.net/20.500.12395/46147 | |
| dc.identifier.volume | 21 | en_US |
| dc.language.iso | en | en_US |
| dc.publisher | Selçuk Üniversitesi | en_US |
| dc.relation.ispartof | Selcuk University Journal of Engineering Sciences | en_US |
| dc.relation.publicationcategory | Makale - Uluslararası Hakemli Dergi - Başka Kurum Yazarı | en_US |
| dc.rights | info:eu-repo/semantics/openAccess | en_US |
| dc.selcuk | 20240510_oaig | en_US |
| dc.subject | Data Mining | en_US |
| dc.subject | Machine Learning | en_US |
| dc.subject | PMML | en_US |
| dc.subject | Software Development | en_US |
| dc.subject | Web Service | en_US |
| dc.title | Performance Comparison of Model Storage Formats for Deploying Data Mining Models | en_US |
| dc.type | Article | en_US |
