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