Search academic texts

Information × Registration Number 0224U033228, (0121U110545) , R & D reports Title Computer technologies, systems, components: modeling and software popup.stage_title Комп’ютерні технології, системи, компоненти: моделювання та програмне забезпечення Head Hunchenko Yurii О., Доктор технічних наук Registration Date 21-12-2024 Organization Odesa I.I. Mechnikov National University popup.description1 The purpose of the work is to increase the efficiency and quality characteristics of intellectual, specialized and computing systems and their components by improving the methods of modeling, analysis and verification of their functioning; improving the software; modeling the proposed technologies popup.description2  The object of the study is processes, methods, models and software of computing systems. A nonlinear regression model with three predictors is proposed for early assessment of the LOC metric of open source applications based on the four-variable Box-Cox transformation, which, in comparison with other models, showed a higher multiple coefficient of determination, a smaller value of the average relative error, and a smaller width of confidence intervals and prediction intervals. A method for early assessment of LOC of open source applications based on a 4-dimensional normalizing transformation is proposed, which allowed to increase the reliability of early assessment of LOC in relations. For the first time, the confidence and prediction intervals of nonlinear regression are constructed using the normalizing Box-Cox transformation, which is based on multiple nonlinear regression analysis using multivariate normalizing transformations, which made it possible to take into account the correlation between dependent and independent variables in the case of non-Gaussian data. The mathematical model in the form of a ten-dimensional prediction ellipsoid for normalized data using the multivariate Box-Cox transformation has been improved. The use of the test (Mardia test) is proposed to check the deviation of the multivariate data distribution from normal. A new intelligent information-measuring system for detecting and analyzing a specific surface by measuring high-resolution electromagnetic waves and infrared radiation has been proposed. A comprehensive review of methods for reducing errors in numerical calculations, which are used to increase the accuracy of mathematical models, has been conducted. This allows to increase the accuracy and reliability of mathematical models for numerical methods and their application in various industries Product Description popup.authors Epik Maryna O. Hunchenko Yurii O. Zui Oksana M. Kamienieva Alla V. Korenkova Hanna V. Kudin Oleh O. Laryn Dmytro H. Martynovych Larysa Ya. Mykhailenko Vladyslav S. Pasenchenko Thomas O. Prykhodko Serhii B. Stukalov Serhii A. Sharipova Ilnara V. Shuhailo Yurii B. popup.nrat_date 2024-12-21 Close