Prediction commercial and cordwood volume of broadleaves using Artificial Neural Networks (Case study: Gorazbon distric of Kheyrood forest, Nowshahr)

Decision-making in natural resources often leads to complexities beyond the statistical empirical methods,therefore we need new solutions than algorithmic methods. Artificial neural networks (ANN) technology mimics the human brain in the process of problem solving.The aim ofthis studywas to predict the commercial volume and cordwood volume using this technique (Artificial Neural Network). For this purpose, 367 marked trees in the experimental and educational forest of Kheyrood were selected. Some factors including diameter at breast height, diameter at stump, stump height, total height, topographic factors (slope, aspect and elevation), species, tree situation and minimum median diameter of last log were measured. The factors were considered as input network. Multi-layer Perceptron network (MLP) was used for modeling. The result showed that Multi-layer Perceptron network (with the 0/94 and 0/71 R2, and 0/233 RMSE) has acceptable accuracy to predict the commercial and cordwood volume.