Designing The Most Suitable Hybrid Model for Forecasting The Future Price of Saffron in The Agricultural Commodity Exchange

Saffron, as the most expensive agricultural and medicinal product in the world, has a special place in buying and selling related to the Iranian agricultural commodity bourse. The saffron commodity bourse, like any other market, has always faced limitations and structural problems, and most of these problems are related to its price fluctuations. Price forecasting using appropriate models can be a great help in reducing the price risk of futures market of saffron. The main question is that in in confronting with various price forecasting methods, which methods should be chosen to forecast the future price of saffron? The purpose of this study is to design the most appropriate hybrid model for forecasting the future price of Negin saffron in the agricultural commodity bourse, which consists of a set of nonlinear models of genetic algorithm, deep neural network, random forest, support vector machine and Monte Carlo method. In this hybrid model, genetic algorithm is used to determine the optimal lag of price time series, deep neural network, random forest model and support vector machine are used to forecast the price time series, and Monte Carlo method is used to simulate the most probable price probability. The results of this study showed that the forecasting accuracy of the hybrid model of "Genetic Algorithm-Deep Neural Network-Monte Carlo" is higher than the two models of "Genetic Algorithm-Random Forest-Monte Carlo" and "Genetic Algorithm-Support Vector Machine -Monte Carlo". Therefore, using a deep neural network and calculating the most probable price probability by the Monte Carlo method, provide the most accurate saffron price prediction with a high degree of reliability and minimal risk. Thus, it is suggested that that the management of the commodity bourse, stock market participants and researchers can use the advantages of this proposed model in forecasting the price of agricultural products.