Отрывок: A symmetric piecewise-linear decision tree is a piecewise-linear decision tree that has all the levels of the tree completely filled, all nodes on the same level have the same splitting rule and common left linear coefficient and right linear coefficient. Such kind of trees are simpler than general piecewise-linear decision trees and can be considered as a method of regularization. They also have an increased prediction speed, having a greater degree of utilization of modern CPU...
Полная запись метаданных
Поле DC | Значение | Язык |
---|---|---|
dc.contributor.author | Gurianov A. | ru |
dc.coverage.spatial | Decision trees | ru |
dc.coverage.spatial | gradient boosting | ru |
dc.coverage.spatial | machine learning | ru |
dc.coverage.spatial | piecewise-linear decision trees | ru |
dc.coverage.spatial | градиентное усиление | ru |
dc.coverage.spatial | деревья решений | ru |
dc.coverage.spatial | кусочно-линейные деревья решений | ru |
dc.coverage.spatial | машинное обучение | ru |
dc.creator | Gurianov A. | ru |
dc.date.issued | 2022 | ru |
dc.identifier | RU\НТБ СГАУ\496000 | ru |
dc.identifier.citation | Gurianov, A. Oblivious piecewise-linear decision trees / A. Gurianov // Информационные технологии и нанотехнологии (ИТНТ-2022) : сб. тр. по материалам VIII Междунар. конф. и молодеж. шк. (г. Самара, 23 - 27 мая) : в 5 т. / М-во науки и образования Рос. Федерации, Самар. нац. исслед. ун-т им. С. П. Королева (Самар. ун-т), Ин-т систем обраб. изобр. РАН - фил. ФНИЦ "Кристаллография и фотоника" РАН. - Самара : Изд-во Самар. ун-та, 2022Т. 5: Науки о данных / под ред. А. В. Куприянова. - 2022. - С. 053462. | ru |
dc.description.abstract | Gradient boosting ensembles of decision trees are a very popular type of machine learning model, with several popular implementations. Some of those implementations utilize symmetric decision trees - decision trees of a specific structure that improve regularization and speed predictions. Over the last years, several research works have been published related to piecewise-linear decision trees and their utilization in gradient boosting ensembles. In this paper, symmetric piecewise-linear decision trees were introduced, and it was shown that it ispossible to efficiently train such trees in gradient boosting ensembles. It was shown that such symmetric piecewise-lineardecision trees have significantly faster prediction time compared to regular decision trees and piecewise-linear decision trees ofsimilar depths and that ensembles of symmetric piecewise-linear decision trees achieve competitive quality on open datasets. | ru |
dc.language.iso | eng | ru |
dc.relation.ispartof | Информационные технологии и нанотехнологии (ИТНТ-2022) : сб. тр. по материалам VIII Междунар. конф. и молодеж. шк. (г. Самара, 23 - 27 мая) : в 5 т. - | ru |
dc.source | Информационные технологии и нанотехнологии (ИТНТ-2022). - Т. 5 : Науки о данных | ru |
dc.title | Oblivious piecewise-linear decision trees | ru |
dc.type | Text | ru |
dc.citation.spage | 053462 | ru |
dc.citation.volume | 5 | ru |
dc.textpart | A symmetric piecewise-linear decision tree is a piecewise-linear decision tree that has all the levels of the tree completely filled, all nodes on the same level have the same splitting rule and common left linear coefficient and right linear coefficient. Such kind of trees are simpler than general piecewise-linear decision trees and can be considered as a method of regularization. They also have an increased prediction speed, having a greater degree of utilization of modern CPU... | - |
Располагается в коллекциях: | Информационные технологии и нанотехнологии |
Файлы этого ресурса:
Файл | Размер | Формат | |
---|---|---|---|
ИТНТ-2022. Том 5. Науки о данных/978-5-7883-1793-9_2022-053462.pdf | 740.23 kB | Adobe PDF | Просмотреть/Открыть |
Показать базовое описание ресурса
Просмотр статистики
Поделиться:
Все ресурсы в архиве электронных ресурсов защищены авторским правом, все права сохранены.