Optimized High Performance Deep Neural Network for Crop Recommendation

Abstract

Crop productivity is most important in the agriculture sector and is impacted by various climate and chemical factors. A huge number of losses bear by farmers just because of these two factors. The climate factors are unpredictable, and human has no control over them but the chemical factor can be controlled by automated techniques. Several solutions have been proposed by research to address such issues. But this paper is concerned to address the issue of crop recommendation based on chemical and climate conditions. In this paper, a grey wolf optimization-based deep learning approach is proposed to suggest better crops based on chemical and climate conditions. This paper considers different chemical factors such as ph, nitrogen, phosphorus and potassium and various climate factors such as rainfall, temperature, and humidity to suggest crops to farmers. The complete approach is laid down in different folds: Firstly, the high-performance Convolution neural network is used to extract important features and classify them and afterward, the grey wolf optimization is used to optimize the feature to suggest a better crop based on different factors.