PREDICTING THE PRICES OF FOREST ENERGY RESOURCES WITH THE USE OF ARTIFICIAL NEURAL NETWORKS (ANNs). THE CASE OF CONIFER FUEL WOOD IN GREECE

The successive oil crises witnessed from the 70's to the present, have led to an increase in the price of oil and uncertainty regarding its availability. Simultaneously, however, the significance of renewable energy sources is also increasing and a global interest has been observed concerning wood as an energy source (quality, quantity, prices). Based on the above, forecasts can provide significant information and define the future structure of the market for fuel wood, because they constitute a decision-making basis for the primary sector, in determining the selling price of fuel wood. This paper is an initial effort to apply an Artificial Neural Network (ANN) in order to assess the future prices of conifer fuel wood in Greece. The used data refer to the period 1964-2005 and portray the average annual prices of all state forest farms that auction conifer fuel wood and come from the production with self-supervision of state forests. The neural network used in this case uses the quantities of conifer fuel wood auctioned during the period 1964-2005 as input values and has the ability to estimate future selling prices with great accuracy (R-2 congruent to 90%).