Применение информационных технологий в эконометрических исследованиях

Abstract

Проводится сравнительный анализ возможностей автоматизации регрессионного анализа средствами пакетов прикладных программ MS Excel и MatLab. Рассматриваются вопросы выбора на стадии спецификации эконометрической модели вида аналитической зависимости с помощью встроенных графических средств системы MatLab.

Проводиться порівняльний аналіз можливостей автоматизації регресійного аналізу засобами пакетів прикладних програм MS Excel та MatLab. Розглядаються питання вибору на стадії специфікації економетричної моделі виду аналітичної залежності за допомогою вбудованих графічних засобів системи MatLab.

The article deals with the comparative analysis of the performance capabilities of the regression analysis automation by means of MS Excel and MatLab. It considers the problems arising when the multi-factor linear regression analysis with the use of MS Excel is carried out; and the problems connected with the solution of the high order linear equation systems and nonlinear econometric models construction. The advantages of the MatLab system are revealed while solving the analogue tasks. The particular example gives the description of methodology of the choice of the type of analytical dependence at the stage of econometric model specification with a help of built-in graphical tools of the MatLab system. It is shown how the results of such choice can be used in the MatLab system during the construction of linear and nonlinear models based on Statistics Toolbox instruments. The construction of the approximation model was carried out both by means of MS Excel and MatLab using the polyfit and nlinfit functions. The quality of the regression equation obtained was estimated by the coefficient of determination value. The results of the experiment conducted show that the use of the MatLab system in econometric research can considerably simplify the process of carrying out the regression analysis.