A Cognitive IoT Learning Models for Agro Climatic Estimation Aiding Farmers in Decision making

climate change continues to be an impact for every nation’s agricultural system, forecasting it is regarded as one of the most significant economic factors. For farmers to survive the increasing frequency of extreme weather events that have a detrimental effect on agricultural production, climate data and services are essential. Weather forecasts are essential for agricultural resource management because they help farmers prepare ahead of time and safeguard their crops from natural calamities. Furthermore, climate data has been fuelled by global warming, resulting in unexpected hurricanes that have even harmed agriculture’s production roots. These days, the daily forecasting of weather variables, such as rainfall, maximum temperature, and humidity, is primarily done using artificial intelligence, machine learning, and deep learning approaches. The current climate condition models require more innovation in terms of high performance and computational complexity. This study suggests Harris Hawk Optimised deep learning network and ensemble residual Long Short-term memory (R-LSTM) for climatic condition prediction that supports an improvement in crop-yield output. The climate parameter is used to train the proposed model, which is then assessed using the several state-of-the-art learning techniques and performance metrics like accuracy, precision, recall, specificity, and F1-score. The results show that the suggested model has a 97.3% accuracy rate, a 96.9% precision rate, a 96.6% recall rate, and a 97.4% F1-score. The results of the current study show that the suggested model is a very good choice for predicting climate change. By increasing crop output productivity, this in turn significantly contributes to raising farmers’ standard of living.