Сlustering method for quality verification of experts’ Unified State Examination checking
Автор: Shcherbinin Timofey A., Grankov Mikhail V.
Журнал: Вестник Донского государственного технического университета @vestnik-donstu
Рубрика: Информатика, вычислительная техника и управление
Статья в выпуске: 3 (86) т.16, 2016 года.
Бесплатный доступ
The possibilities of using the fuzzy clustering method in the automated system of evaluation working efficiency of the experts’ validation status on the assessment of the state final certification in the form of the Unified State Exam (USE) are considered. Various parameters of experts’ evaluation and their calculation formula are presented. The parameters are divided into the following groups: basic statistical parameters, advanced statistical parameters, normalized parameters, and quality parameters. The application of the fuzzy clustering method using a weighted Euclidean metric is illustrated by the case of the assessment of the experts’ work in the subject commission on mathematics (profession-oriented level). The experiments are performed with and without noises with the subsequent conclusions of the impact of noises on the results. Besides, the experimental results are presented in tables and graphs with partitioning of experts into clusters.
Fuzzy logic, cluster analysis, unified state exam, evaluation, noise in the data, term-sets
Короткий адрес: https://sciup.org/14250217
IDR: 14250217 | DOI: 10.12737/20222