جستجوی مقالات مرتبط با کلیدواژه « Artificial Neural Network » در نشریات گروه « اقتصاد کشاورزی »

Lavender with the scientific name Lavandula stricta Del is a perennial medicinal plant with a height of about half a meter that grows in different regions of Iran. Drying is one of the oldest methods of preserving materials. The use of neural networks can be used in the design and selection of optimal working conditions and dryer control. In this study, various parameters of drying, evaluation of mathematical models to determine the best model, evaluation of different topologies of MLP artificial neural network to determine the best network for lavender plant with microwave dryer with power range of 100-1000 watts and The frequency of 2450 MHz is provided in four power levels of 300, 500, 700 and 900 watts. MLP artificial neural network was used to predict the relationship between drying kinetic parameters (moisture ratio and drying rate) and efficiency of energy consumption with changes in microwave power consumption using Statistica software. Among the fitted models, the Midili model was chosen as the best model according to R 2, χ 2 and RMSE criteria. Microwave power levels had an effect on drying time, with drying times of 3 minutes for 900 W power and 11 minutes for 300 W power. In order to predict drying kinetic parameters and energy consumption efficiency, MLP network with one input and three outputs was successfully used. The results generally showed that the MLP artificial neural network is a very powerful tool in predicting drying kinetic parameters and energy efficiency of lavender medicinal plant based on microwave power consumption values.

In this research, we try to predict the yield of rainfed barley in Khorasan provinces using climatic parameters and two methods of artificial nervous netwework (Ann) and fuzzy neural system (Anfis). Calculations were performed with MATLAB software and then the statistical indices of correlation coefficient (R2), root mean square error (RMSE) and full mean error (MAE) were used to evaluate the performance of the models. Last year's yield and rainfall had an effective role in reducing prediction error and increasing correlation coefficient in both Ann and Anfis methods. Last year's yield and evapotranspiration made the Anfis method more accurate than the Ann method. The results of both Anfis and Ann methods for model L inputs, which included rainfall, relative humidity and last year's yield, showed that this model achieved the highest accuracy among the input models. However, in the Anfis method for model E inputs, which included evapotranspiration, rainfall, relative humidity and minimum temperature, the results showed that it was more accurate than the Ann method. The greatest difference in accuracy in estimating yield between the two Anfis and Ann methods was observed with R inputs model, which includes moisture inputs, Dew point temperatures and maximum temperatures. The presence of radiation parameters at the inputs reduced the accuracy of yield estimation in both methods. Overall, the Anfis method was more accurate in estimating yield than Ann.

The purpose of this study is to investigate the effect of imports in the agricultural sector on the demand of agricultural labor (employment) in Iran for the period 1978-2019 and forecast employment in this sector for the outlook 2026. In the present study, labor demand in the agricultural sector is a function of wages, value added in the agricultural sector, capital stock and imports. To estimate the aggregate relationship and short-term and long-term dynamics, vector autoregression patterns and vector error correction model were used. Also, the neural network designed due to greater efficiency than other models was used to predict employment in the agricultural sector of Iran for the years 2020 to 2026. The results showed that in the long run, value added in the agricultural sector has a positive effect and capital stock, wages and imports have a negative effect on employment in the agricultural sector. The results also showed that the trend of reducing the share of employment in the agricultural sector of Iran in relation to the total economy until 2026, will continue as in previous decades and from an annual growth of 1.03 percent to 0.77 percent. Ultimately, government policies must be implemented with the aim of increasing the value added of the agricultural sector in order to provide new jobs, food security and self-sufficiency.

One of the requirements of planning for the future is predict the behavior of economic variables. Since wheat is a strategic commodity for our country, forecasting its price is very important. In previous studies in Iran, researchers have used quantitative models to forecasting the price of wheat and they have not used qualitative models. But in present research, we use both of them. The annual data for period of 1976-2014 are included. The results of the study indicat that RMSE criterion for quantitative models such as ARMA, EGARCH and ANN are 37625.68, 39373.91 and 24258.073, respectively. On the other hand, the average percent difference between the forecasting of ANN and Delphi method is 0.08. So, the results show that prediction error of the neural network model compared to other methods is smaller and in prediction of future price compared with qualitative methods (Delphi model) is slightly different. It indicates the importance of using qualitative methods beside quantitative methods for forecasting economic variables.

The importance of economic variables forecasting for policy makers and planners and firms is obvious. Therefore, in recent decades, several models have been developed. In the present study, the Oilcake import quantity predicted by using econometric Model and artificial neural network for 1400-1394. For this purpose, for predict and education network, the data of the 88-1348 and for test the accuracy of forecasts, the data of the 1393-1389 were used. The results showed that forward neural network compared to ARIMA and exponential smoothing method with fewer errors and better performance for predicting of Oilcake import. The results show that the Oilcake imports in 1394 compared to last year, a 32% increase. Therefore, it is necessary to reduce the amount of imported and domestic needs, be protectionist policies. Also, to control of the steep rise in Oilcake import, policy makers and planners would be necessary programs apply, including increased tariffs. The grants for the production and supply companies in the conversion process to oilcake can be used to increase the productivity and competitiveness of these products to be effective.

In spite of large emphasis on agriculture and food security in the upstream documentation and national development plans, importation of agricultural and food products has had an upward trend mainly influenced by various variables within last three decades. Imports of these products have increased of nearly 5 million tons in 1981 to more than 21 million tons in 2012. Aims of present study is identifying and examining the effective factors on Iranian agricultural import demand for the period 1991-2012 using combination of exploratory data analysis methods (EDA) with Artificial Neural Network. For this purpose, information of eleven goods consist of Wheat, barley, rice, corn, Soybean meal, oil (soybean and sunflower), sugar, eggs, milk, poultry and beef meat have been used. The result of modeling shows that nonoil GDP (29.5%), oil incomes (27.7%), tariff rate (16.5%) and amount of domestic production (14.6%) have the greatest effect on import demand. Also the results revealed that relative price of agricultural products imports (6.6%) and lag of import demand (5.1%) have the lowest effect on import demand.

The main purpose of this study is forecasting the production indexes of sugar beet (land, production and price) in Iran. For this purpose, Models applied to forecast are Auto Regressive (ARMA and ARIMA), Single and Double Exponential Smoothing, Harmonic, Artificial Neural Network and ARCH. The results of Durbin-Watson tests were revealed that land, production and price of sugar beet series are non stochastic. According to the lowest forecasting error criterion, Neural Network, ARMA and ARIMA are the best models for forecast land, production and price of sugar beet series respectively. Also, results of prediction show that all of these indexes will have a increasing trend during the period 2013- 2020, while the increasing trend of land and production are more smoother than price trend.

Paying attention to the relative stability of prices and forecasting prices could play an important role in controlling the instability of prices and ultimately reduce market risk. Comparison of different methods is important in the forecasting issues. In this study Wholesale and Retail Prices (Weekly Prices) of Rainbow Trout Fish will forecast with contrasting between forecasting power of the ARMA method and Artificial Neural Network method and choice the better one. In this study the Feed-forward network that is one of Back Propagation networks is used. The using data are from first week of farvardin 1388 to last week of shahrivar 1390. Before usage of predict methods، the random or non-randomized nature of the data were examined. Both of Price Series are predictable and non-randomized based on Random tests of Wald-Wolfowitz، Wallis-Moore and Durbin –Watson. Series are stationary in levels based on data stationary test (Dicky-Fuller augmented). Results of Forecasting show that in the model ARMA compared with artificial neural network (ANN) error rate is less based on four criteria forecast accuracy evaluating. Then it has higher power in forecasting the price of Rainbow Trout Fish. In the ANN model 80% of data for training network and 20% for testing network have been used. The results of accuracy equality test of the two methods (MGN) shows the ARMA model is also better than the neural network model in forecasting retail and wholesale prices significantly.

This study contributes to reduce bread waste in the production process by determining effective factors that distinguish high bread wast bakers from low bread waste bakers using 250 bakeries over Mashhad in the year 2010. The discriminant analysis was used to predict the study bakers into one class of high or low waste groups. Results indicate that among discriminators, bakery status, bread waste price, dough fermentation time, daily consumption of flour, quality of produced bread, percentage of wet gluten, moisture percentage, quality of flour and maintenance costs have the highest share in distingushing between high and low bread waste bakers. Predicting bakers based on their bread waste was considered as a suitable instrument in order to identify effective actions for reducing bread waste. In this study, classification accuracy of discriminant analysis (DA) and DA-based artificial neural network indicates high accuracy of class prediction at training and testing data with DA-based artificial neural network model. Ultimately, based on the results, a number of applicable and executive recommendations to decrese bread waste in the production process were presented.

The main object of this study is to predict electricity consumption of agricultural sector in Iran. To get the objective, time series method of Auto-Regressive Moving Average (ARMA) and artificial neural networks (ANN) were used. Annual data for period of 1967 to 2008 was used. The Mean Absolute Percent Error (MAPE), Root of Mean of Squared Error (RMSE) and Mean Absolute Error (MAE) criteria were used for comparing the ability of different forecasting methods. As the result showed Feed Forward artificial neural network with back proportion algorithm can predict electricity consumption with MAPE equal to 1.02%, while the corresponding value for time series model obtained 1.13 percent. Other criteria also revealed the same result, so, ANN is expected to predict electricity consumption more precise than ARMA model. Therefore, energy ministry may use ANN in future predictions.