Отрывок: 5) then Hiτ� = Change,else Hiτ� = Anomaly. The algorithm for finding anomalies in the time series begins with the expert entering patterns of anomalies. Anomaly pattern is a sequence of numbers of situations that precede anomalies. The last number in the sequence is the number of the abnormal situation. In the work, the algorithm uses one pair of parameters: either a fuzzy label - a fuzzy trend, or measures of entropy by membership function and by a fuzzy trend. Regardless of the choice o...
Полная запись метаданных
Поле DC Значение Язык
dc.contributor.authorTimina, I.A.-
dc.contributor.authorEgov, E.N.-
dc.contributor.authorRomanov, A.A.-
dc.date.accessioned2018-05-18 14:46:38-
dc.date.available2018-05-18 14:46:38-
dc.date.issued2018-
dc.identifierDspace\SGAU\20180518\69522ru
dc.identifier.citationTimina I.A. Application of the anomaly pattern in forecasting time series of project activity metrics/I.A.Timina, E.N.Egov , A.A. Romanov// Сборник трудов IV международной конференции и молодежной школы «Информационные технологии и нанотехнологии» (ИТНТ-2018) - Самара: Новая техника, 2018. - С. 1794-1799ru
dc.identifier.urihttp://repo.ssau.ru/handle/Informacionnye-tehnologii-i-nanotehnologii/Application-of-the-anomaly-pattern-in-forecasting-time-series-of-project-activity-metrics-69522-
dc.description.abstractThis article describes the method of application of an anomaly template of time series of project metrics based on entropy. The analysis of project activity metrics is described. The forecasting algorithm based on the fuzzy trends of time series of indicators was developed and implemented. The formula for the entropy measure for the fuzzy time series is determined. The algorithm uses the dependence of the forecast on the measures of entropy. The hypothesis of trend stability is used for forecasting. Experiments based on this approach are presented.ru
dc.description.sponsorshipThe article was supported by the Russian Foundation for Basic Research (grant No. 16-47-732070).ru
dc.language.isoenru
dc.publisherНовая техникаru
dc.relation.ispartofseries3;241-
dc.subjectfuzzy time seriesru
dc.subjectfuzzy trendru
dc.subjectentropyru
dc.subjecthypothesesru
dc.subjectforecastingru
dc.titleApplication of the anomaly pattern in forecasting time series of project activity metricsru
dc.typeArticleru
dc.textpart5) then Hiτ� = Change,else Hiτ� = Anomaly. The algorithm for finding anomalies in the time series begins with the expert entering patterns of anomalies. Anomaly pattern is a sequence of numbers of situations that precede anomalies. The last number in the sequence is the number of the abnormal situation. In the work, the algorithm uses one pair of parameters: either a fuzzy label - a fuzzy trend, or measures of entropy by membership function and by a fuzzy trend. Regardless of the choice o...-
Располагается в коллекциях: Информационные технологии и нанотехнологии

Файлы этого ресурса:
Файл Описание Размер Формат  
paper_241.pdfОсновная статья.404.64 kBAdobe PDFПросмотреть/Открыть
Показать базовое описание ресурса Просмотр статистики
Поделиться:



Все ресурсы в архиве электронных ресурсов защищены авторским правом, все права сохранены.