Исследование информационной значимости параметров плана медико-биологического эксперимента
Вантажиться...
Дата
2018
Автори
ORCID
DOI
10.20998/2413-4295.2018.09.23
Науковий ступінь
Рівень дисертації
Шифр та назва спеціальності
Рада захисту
Установа захисту
Науковий керівник
Члени комітету
Видавець
НТУ "ХПИ"
Анотація
В работе проведено исследование влияния параметров плана на количество ожидаемой информации базовых моделей медико-биологических экспериментов. Приведены базовые модели планирования медико-биологических экспериментов, полученные на основе классической информационной теории. С использованием разложения Тейлора получены модели, позволяющие независимо оценивать влияние абсолютного и относительного изменения каждого из параметров плана на общее информационное содержание эксперимента. По результатам исследований получены количественные и качественные оценки каждого из параметров плана на количество ожидаемой информации для параметрической, компонент дисперсий и регрессионной моделей планирования эксперимента.
In the work the influence of plan parameters on the amount of expected information of basic models of medical and biological experiments was carried out. The necessity and formalization of the task of investigating the informational significance of the experimental parameters taking into account the statistical properties of biological objects was substantiated. The basic models of planning medical and biological experiments, obtained on the basis of the classical information theory, were given. With the use of the Taylor expansion, models have been obtained that make it possible independently to estimate the influence of the absolute and relative changes of each plan parameter on the general information content of the experiment. Based on the results of the research, quantitative and qualitative estimates of each of the plan parameters for the amount of expected information for the parametric, variance components and regression models of the experiment planning were obtained. As a result of the work, the information theory of measurements was developed for the planning of medical and biological measurement experiments, in which a functional relationship between the parameters of the experimental design and the expected amount of information was disclosed. A mathematical model of the effect of changes in the experimental plan parameters on changes in the expected amount of information has been developed, which made it possible to rank the parameters of the plan according to the degree of diminishing their influence on the amount of experimental information. Such a ranking makes it possible to correlate the values of the plan parameters with conditions of significant limitations on the volume of measurement results. The possibility of an optimal choice of an experiment design model (parametric or random model) is shown, using the expected amount of information as an objective function and an a priori known quantitative effects of factorial influence.
In the work the influence of plan parameters on the amount of expected information of basic models of medical and biological experiments was carried out. The necessity and formalization of the task of investigating the informational significance of the experimental parameters taking into account the statistical properties of biological objects was substantiated. The basic models of planning medical and biological experiments, obtained on the basis of the classical information theory, were given. With the use of the Taylor expansion, models have been obtained that make it possible independently to estimate the influence of the absolute and relative changes of each plan parameter on the general information content of the experiment. Based on the results of the research, quantitative and qualitative estimates of each of the plan parameters for the amount of expected information for the parametric, variance components and regression models of the experiment planning were obtained. As a result of the work, the information theory of measurements was developed for the planning of medical and biological measurement experiments, in which a functional relationship between the parameters of the experimental design and the expected amount of information was disclosed. A mathematical model of the effect of changes in the experimental plan parameters on changes in the expected amount of information has been developed, which made it possible to rank the parameters of the plan according to the degree of diminishing their influence on the amount of experimental information. Such a ranking makes it possible to correlate the values of the plan parameters with conditions of significant limitations on the volume of measurement results. The possibility of an optimal choice of an experiment design model (parametric or random model) is shown, using the expected amount of information as an objective function and an a priori known quantitative effects of factorial influence.
Опис
Ключові слова
планирование эксперимента, параметрическая модель, регрессионная модель, модель компонент дисперсий, уровень значимости, параметр нецентральности, информационный анализ, experiment planning, parametric model, regression model
Бібліографічний опис
Томашевский Р. С. Исследование информационной значимости параметров плана медико-биологического эксперимента / Р. С. Томашевский // Вісник Нац. техн. ун-ту "ХПІ" : зб. наук. пр. Сер. : Нові рішення в сучасних технологіях = Bulletin of the National Technical University "KhPI" : coll. works. Ser. : New solutions in modern technologies. – Харків : НТУ "ХПІ", 2018. – № 9 (1285). – С. 159-165.