Advanced smart farming system based multi-anchor space-aware temporal convolutional neural networks in internet-of-things

Agriculture is an important to the economic growth of a country. Farmers possess until recently employed standard farming methods. Accurate farming helps boost output by accurately identifying the actions that must be taken at the right time. Precision farming includes forecasting the weather, evaluating the soil, suggesting crops to grow, and figuring out the fertilizer the crops require. In this paper, an Advanced Smart Farming System based Multi-anchor Space-aware Temporal Convolutional Neural Networks in Internet-of-Things (ASFS-MSTCNN-IoT) is proposed. Initially, the input data is taken from Indian Agriculture Dataset. Then, the input data is pre-processed utilizingCompact Maximal Correntropy-derived Error State Kalman Filter (CMCESKF)which is used to remove the outliers from the input data. The pre-processed data are given into Deep Kernel Principal Component Analysis (DKPCA)which reduces the high dimensionality of the data. Generally, MSTCNN does not show any adaption of optimization methods for finding the optimal parameters to ensure exactforecastof crop yield. Black-Winged Kite Algorithm (BWKA) is proposed in this work to optimize the weight parameter of MSTCNN classifier, which predicts the crop yield precisely. The ASFS-MSTCNN-IoT approach is implemented and analyzed with the help of performance metrics like Mean Absolute Error (MAE), Mean Square Error (MSE), Mean Absolute Percentage Error (MAPE), R² and Root Mean Square Error (RMSE) is evaluated. Performance of the ASFS-MSTCNN-IoT approach attains17.85%, 25.82%, 32.64% lower Mean Absolute Error, 25.43%, 19.94%, 31.68% lower Mean Absolute Percentage Error and 18.59%, 25.64% and 31.89% higher R² with existing methods respectively.