Red fox optimization with ensemble recurrent neural network for crop recommendation and yield prediction model

Precision agriculture concentrates on monitoring (sensing technologies), management information system, variable rate technologies, and responses to inter- and intravariability in cropping systems. The advantages of precision agriculture involve improving crop productivity and crop quality with minimum environmental impact. Crop yield prediction (CYP) is one of the challenging tasks in agriculture, which mainly depends upon soil, meteorological, environmental, and crop-related variables. On the other hand, farmers usually follow conventional farming patterns to decide on crops to be cultivated in a field. An automated crop recommendation system is required to assist farmers in making informed decisions prior to crop cultivation. Deep learning (DL) and Machine learning (ML) methods provide a practical approach for enhanced crop production and yield prediction using different features. Therefore, this study focuses on the design of Red Fox Optimization with Ensemble Recurrent Neural Network for Crop Recommendation and Yield Prediction (RFOERNN-CRYP) model. The presented RFOERNN-CRYP model follows an ensemble learning process, which makes use of three different DL models (namely long short-term memory (LSTM), bidirectional LSTM (BiLSTM), and gated recurrent unit (GRU)) for accomplishing enhanced prediction performance compared to the individual classifier models. Moreover, the RFO algorithm is applied for the hyperparameter selection of the three DL models to improve the overall performance, showing the novelty of the work. The experimental validation of the RFOERNN-CRYP technique is validated on crop recommendation and yield prediction datasets from the Kaggle repository. The experimental outcomes showed that the proposed model outperforms the other recent approaches regarding several measures. The presented RFOERNN-CRYP technique assists farmers in the decision-making process using different agro parameters.