APPLICATION OF DISCRIMINANT ANALYSIS AND NEUTRAL NETWORKS TO FORECASTING THE FINANCIAL STANDING OF FARMS WITH CONSIDERATION THE INFLUENCE OF THE TIME

Abstract

The aim of the research was to determinate a linear discriminant function and neural network that could be applied for classifi cation of farms. The result of classifi cation was forecasting fi nancial situation of them, based on set of many variables, which including fi nancial indicators. Models were built separately for each year, but they were verifi ed in the rest. The additional aims were to determine the set of indicators with large forecasting ability, and to compare two classifi cation methods – linear discriminant function and neural network.