Multitudinous Transfer Learning Based Yield Prediction of Paddy by Enhancing Sustainable Smart Agriculture

Smart agriculture, leveraging deep learning to balance global food demand and ecological preservation, offers hope for modern farming. Deep learning algorithms detect pests and diseases early, reducing yield losses and minimizing overall pesticide use. Precision irrigation conserves water and nurtures soil life through data-driven techniques. Yield prediction, incorporating historical and current data, aids in wise decision-making, while ecosystem monitoring promotes biodiversity and resilience to extreme weather. Deep learning contributes to soil health improvement and carbon sequestration to address climate change. The study compares the effectiveness of bi-directional LSTM (Bi-LSTM) and Multitudinous Transfer Learning (MTL), The MTL model achieved higher accuracy (0.90) compared to the Bi-LSTM model (0.88), with MTL proving to deliver superior results, enhancing the fusion of technology and sustainability.