Уточнення методу обробки статистичних даних
Вантажиться...
Дата
2014
Автори
Горошко, Андрій Володимирович
Goroshko, A.V.
Ройзман, Вілен Петрович
Royzman, V.P.
Назва журналу
Номер ISSN
Назва тому
Видавець
Національна академія Державної прикордонної служби України імені Богдана Хмельницького
Анотація
Проаналізовано методи статистичної обробки емпіричних даних в процесі вирішення проблем моніторингу стану прикордонної
безпеки, надійності технічних засобів охорони державного кордону та ін. Запропонований метод уточненої обробки статистичних даних, подаючи закон розподілу у вигляді полімодальної суміші унімодальних законів. Метод дозволяє розкрити внутрішню структуру даних з урахуванням можливої полімодальності закону їх розподілу і поряд з обґрунтованим вибором кроку побудови гістограм дає правила роботи з такими даними. Подані результати практичного
застосування методу.
Handling statistical data is an essential component to solving the problems of monitoring the state of border security, assessment and prediction of the sources and nature of the threats and challenges to national interests at the state border of Ukraine, development and study of methods for evaluating the effectiveness of the operational activities of the border agency, techniques calculation needs men and equipment necessary for the effective performance of assigned tasks, study methodologies to assess the efficiency of the management structures of the (sub) state border protection , reliability, technical means of protection of the state border, and others. In contrast to the known method adopted by the authors propose to treat the data, suggesting that the observation made by “emissions” are those of the individual modes and distribution law really is not unimodal and polymodal. The method is presented and processed empirical density distribution as a superposition of features unimodal density distribution (mixing) with the a priori probabilities (weights). The authors propose a method based on decomposition of mixtures approximation of probability density by the method of least squares interpolation or some dotted-set. The paper presents a method of processing refined empirical statistical data representing the distribution law as polymodal mixture unimodal laws. The method can reveal the internal structure of the data, taking into account possible polimodal law and their distribution, along with a reasonable selection step histogram gives the rules for dealing with such data. In subsequent publications, the authors will describe the method for assigning confidence intervals for polymodal distribution laws and their practical application to determine the acceptable values of observational data.
Handling statistical data is an essential component to solving the problems of monitoring the state of border security, assessment and prediction of the sources and nature of the threats and challenges to national interests at the state border of Ukraine, development and study of methods for evaluating the effectiveness of the operational activities of the border agency, techniques calculation needs men and equipment necessary for the effective performance of assigned tasks, study methodologies to assess the efficiency of the management structures of the (sub) state border protection , reliability, technical means of protection of the state border, and others. In contrast to the known method adopted by the authors propose to treat the data, suggesting that the observation made by “emissions” are those of the individual modes and distribution law really is not unimodal and polymodal. The method is presented and processed empirical density distribution as a superposition of features unimodal density distribution (mixing) with the a priori probabilities (weights). The authors propose a method based on decomposition of mixtures approximation of probability density by the method of least squares interpolation or some dotted-set. The paper presents a method of processing refined empirical statistical data representing the distribution law as polymodal mixture unimodal laws. The method can reveal the internal structure of the data, taking into account possible polimodal law and their distribution, along with a reasonable selection step histogram gives the rules for dealing with such data. In subsequent publications, the authors will describe the method for assigning confidence intervals for polymodal distribution laws and their practical application to determine the acceptable values of observational data.
Опис
Ключові слова
статистична обробка, закон розподілу, полімодальність, суміш законів розподілу, statistical processing, distribution law, polymodal, the mixture distribution laws
Бібліографічний опис
Горошко А. В. Уточнення методу обробки статистичних даних / А. В. Горошко, В. П. Ройзман // Збірник наукових праць Національної академії Державної прикордонної служби України. Сер. : Військові та технічні науки. - 2014. - № 1. - С. 304–313.