Prediction of agricultural drying using multi-layer perceptron network, long short term memory network and regression methods

One of the processes used in production fertilizers, which has become an important part agriculture, is drying process. Determination proper parameters both terms product quality and efficiency. Regression methods are to determine process frequently. In this study, addition regression method, machine learning techniques also examined such as artificial neural network, long short term memory method. The data obtained from a commercial organomineral fertilizer consisting mixture 5% nitrogen 10% phosphorus at 70?C, 75?C, 80?C were for modelling. simulation results models experiments compared. predictions performances each model presented. appropriate It efficiency product. addition, selection plays role obtaining successful simulations. As result, it been observed that prediction performance created with network more than others. While regressions efficient modelling existing data, they not predicting. Moreover, enough predict peak pits data.