Socioeconomic Changes Based Climate Training for Agricultural Application Using Deep Learning Model

• Published in: Remote sensing in earth systems sciences (Online) 2024-12, Vol.7 (4), p.399-410
• Main Authors: Sunitha, M., Durairaj, M., Rajalingam, A., Yusoff, Siti Khalidah Mohd, Prasad, S. Hari Chandra, Malluvalasa, S. N. Lakshmi, Kiran, Ajmeera
• Format: Article
• Language: English
• Subjects: Earth and Environmental Science; Earth System Sciences; Geography; Monitoring/Environmental Analysis; Remote Sensing/Photogrammetry
• ISSN: 2520-8195; 2520-8209
• DOI: 10.1007/s41976-024-00132-0

In socioeconomic factor studies, the use of the logit and probit models has the disadvantage of representing random variation in preference for unobserved components and time-linked errors. The agricultural sector in the coastal nations of South-East Asia is facing mounting strain due to the adverse effects of climate change. Anticipating and controlling climate variables is challenging and necessitates expensive fixes. Based on social and economic shifts and the use of a deep learning model to detect climate change, the two main facets of agricultural application have been examined in this study. The suggested model focusses on these two factors because the economics and climate are crucial components of agriculture. Initially, financial flow–based data analysis employing recurrent LSTM spatial neural networks for text data analysis was used to assess the social economic analysis for those countries with minimal agricultural development. The study of crop growth data based on climate change is then done using Boltzmann Markov quantile regression neural networks. In terms of training accuracy, sensitivity specificity, and ROC, a performance study is conducted for crop growth data based on climatic change and agricultural development based on economic data. The proposed approach in this instance produced training accuracy of 79%, specificity of 83%, ROC of 84%, and sensitivity of 80% for economic data–based agricultural development; the suggested method achieved 75% training accuracy, 73% specificity, 80% ROC, and 84% SENSITIVITY for climate change–based crop growth data.