Search academic texts

Information × Registration Number 0211U008035, 0109U006197 , R & D reports Title Optimization of diagnostic methods of multivariate statistical analysis popup.stage_title Head Lebedev Evgen Alexandrovich, Доктор фізико-математичних наук Registration Date 17-03-2011 Organization Taras Shevchenko Kiev university popup.description2 A survey of the main probabilistic and statistical methods of ordering factors and finding of distance between experimental and control samples is presented. Peculiarities of their implementations in various software are also presented. For calculation of the distributions of mathematical statistic such as Gaussian, Student's, and Snedecor-Fisher's ones the fast recurrent schemes are worked out. Based on the Mahalanobis distance and Snedecor-Fisher's distribution the standardized distance between experimental and control samples is constructed. Using results of correlation analysis the subset of factors that provide a maximal value of the distance is constructed. Software for treatment of multivariate statistical data arrays is created by application of the programming language Builder Client/Server Suit. This software stores, corrects and accumulates information and results of statistical analysis; calculates the standardized distance between experimental and control samples; finds the subset of factors that provide a maximal value of the distance. Product Description popup.authors Лебєдєв Євген Олександрович Лукович Ольга Василівна Макушенко І. popup.nrat_date 2020-04-02 Close