Design and Development of Gorilla Optimized Deep Resilient Architecture for Prediction of Agro-Climatic Changes to Increase the Crop–Yield Production

Predicting climatic changes is considered one of the most important economic parameters as it remains a catalyst for the agricultural system of any country. Climatic data and services are crucial for agriculturalists to withstand the rising frequency of strong meteorological conditions, which negatively impact crop yields. Weather forecasts play a key role in managing resources for agricultural operations, enabling farmers to plan and protect their crops from natural disasters. In addition, global warming has fueled climatic unpredictability, creating challenges like hurricanes that damage the foundational roots of agricultural production. In recent times, Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) techniques have been predominantly adopted for daily forecasting climatic conditions, including rainfall, maximum temperature, and humidity. However, the existing models for climatic prediction require improvements in computational complexity and performance. This research article proposes the ensemble Residual Long Short-Term Memory (R-LSTM) along with Artificial Gorilla Troops Optimized Deep Learning Networks (AGTO-DLN) as a solution for climatic condition prediction to boost crop–yield production. Performance metrics for the proposed model examining precision, F1 score, accuracy, specificity, and recall operate through evaluation using 5,04,647 climatic parameters with various advanced learning techniques. The proposed method achieved 97.2% accuracy and 96.9% precision together with 96.5% recall and 96.6% specificity and 97.5% F1 score. The research proves that the proposed model demonstrates high potential for climate forecast in agricultural production environments consequently boosting crop yields to enhance farmer incomes.