Forecasting the profit for the Greek non-metallic sector using a neuro-fuzzy approach (ANFIS)

Changes have occurred in the non-metallic minerals market. The Greek cement and building materials market was expected to be more connected with the technical-constructing market as has already happened in other countries. The profits incurred a reduction due to a decline of building activity, recession of the demand regarding products in the European market, and an intensified competition from abroad. Producers were obliged to absorb the biggest part of the incremental production cost when the cost of the production factors increased, in order not to suffer losses in sales and market shares (FGI, 2000). The financial and economic applications of soft-computing methods have attracted a lot of interest in past years. It had become clear to many observers that soft computing and computer science tools, especially those from the fields of fuzzy systems and neuro-computing were frequently finding relevance. One of the approaches that have been tried to improve the ability of forecasting is the Artificial Neural Networks. The main goal of this study is to model the relationship between the gross profit and some given macro-economic and micro-economic variables. The data has been generated by the economic market system. The analyzed companies were selected to be from the 200 largest manufacturing companies by net sales and between the 15 largest manufacturing companies in the industrial group by total assets. A data set of yearly observations of the profit and sales of these companies and some relevant macroeconomic factors is considered. After the data has been collected, an ANFIS is utilized to capture the relationship between the independent and dependent variables.