Optimizing Artificial Neural Network for Demography Based Crop Recommendation: An Ocean Water Current Inspired Approach in Precision Agriculture

The challenge of optimal crop selection becomes more complex by the dynamic nature of the ever-changing features of land surface type, soil type, climate, water requirements, phosphorus content, Leaf area index (LAI), and other phenological parameters. In this research, Landsat 8 satellite imagery is used to create an optimized crop recommendation system, extending an Artificial Neural Network (ANN) model with an Ocean water current optimizer for prediction. This research work aims to provide farmers with effective crop recommendations while addressing a wide range of environmental and operational factors. The approach relies on a classifier and an optimizer to choose the best crop to grow in a given region, taking into account all of the features of a specific demography. The training time of the ANN model is also reduced by using OWCO optimizer as compared to other metaheuristic models. The various computing techniques such as Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL) and Nature inspired Optimization Algorithms (NIOAs) for smart farming are also analyzed. The proposed approach complements the ANN by also handling dynamic and mutating factors in the crop recommendation system. The ANN model consistently shows better training accuracy as the number of epochs approaches 60. The accuracy of OWCO with ANN is 94.12% which is better than other metaheuristic models.