Page 89 - Contributed Paper Session (CPS) - Volume 1

CPS1119 Fethi Şaban ÖZBEK

The development of national reporting platform
for global SDG indicators in Finland
Fethi Şaban Özbek
Turkish Statistical Institute

Abstract
Forecasting agricultural prices is useful for farmers, policymakers, and
agribusiness industries. In this study, wheat, barley, maize (seed), and raw
cotton, widely sown in arable land of Turkey (57% of total agricultural area
excluding fallow area), were selected for forecasting by using artificial neural
networks (ANN). The data of monthly price from 2009 January to 2018
September were used as sample data (network training) and historical
prediction evaluation (network evaluation). In each forecasting, the input was
the last observation, and the output was the predicted value of agricultural
commodity price. ANN models were effective for forecasting agricultural
commodity prices in that all accuracies were very high; 99.2, 90.9, 99.0, 91.7
for wheat, barley, maize (seed), and raw cotton, respectively. The results of
forecasting models show that the prices of wheat fluctuate between 0.97
TL/KG and 0.93 TL/kg between 2018-October and 2019-December. And the
prices fluctuate between 0.84 TL/KG and 0.81 TL/kg, 0.87 TL/kg and 0.84 TL/kg,
2.26 TL/kg and 2.24 TL/kg for maize (seed), barley and raw cotton, respectively.

Keywords
Agricultural prices; Artificial neural network; Forecasting; Turkey agriculture

1. Introduction
It is well known that forecasting agricultural prices is useful for farmers,
policymakers, and agribusiness industries. Wheat, barley, maize (seed), and
raw cotton, widely sown in arable land of Turkey (57% of total agricultural area
excluding fallow area), were selected for forecasting.
In recent years, machine learning techniques, such as decision trees and
artificial neural networks (ANN), because they are quick, powerful, and flexible
tools for prediction, classification, optimization, and decision support system,
are used increasingly in agriculture. ANN has provided great benefits to the
researchers for forecasting in economics and finance (Zhang et al., 1998; Jha
et al., 2009), particularly in agricultural price forecasting (Li et al., 2010; Jha and
Sinha, 2013). Also, artificial neural network (ANN) has been proposed as an
efficient tool for modelling and forecasting (Ozbek and Fidan, 2008; Li et al.,
2010; Jha and Sinha, 2013; Al-Maqaleh et al., 2016 etc.).

78 | I S I W S C 2 0 1 9