首页|Sun Yat-sen University Reports Findings in Machine Learning (Upscaling net ecosy stem CO2 exchanges in croplands: The application of integrating object-based ima ge analysis and machine learning approaches)
Sun Yat-sen University Reports Findings in Machine Learning (Upscaling net ecosy stem CO2 exchanges in croplands: The application of integrating object-based ima ge analysis and machine learning approaches)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – New research on Machine Learning is th e subject of a report. According to news originating from Guangdong, People’s Re public of China, by NewsRx correspondents, research stated, “Accurately estimati ng the net ecosystem exchange of CO (NEE) in cropland ecosystems is essential fo r understanding the impacts of agricultural practices and climate conditions. Ho wever, significant uncertainties persist in the estimation of regional cropland NEE due to landscape heterogeneity and variations in the efficacy of upscaling m odels.” Our news journalists obtained a quote from the research from Sun Yat-sen Univers ity, “Here, we applied an integrated approach that combined object-based image a nalysis (OBIA) techniques with advanced machine learning (ML) approaches to upsc ale regional cropland NEE. We conducted a thorough evaluation of the upscaling a pproach across four distinct cropland areas characterized by diverse climate con ditions. Our study confirmed that OBIA techniques can efficiently segment cropla nd objects, thereby enhancing the representation and accuracy of characteristics relevant to cropland features. The sequential least squares programming algorit hm, among the three methods used for ML model integration, demonstrated exceptio nal performance in predicting NEE, with an R value exceeding 0.80 across all stu dy areas and peaking at 0.90 in the most successful area. On average, there was an 18 % improvement compared to the poorestperforming ML model an d a 6 % enhancement compared to the best-performing ML model. The upscaled regional products exhibited superior performance in characterizing crop land NEE patterns compared to pixel-based products. Additionally, we utilized th e SHapley Additive exPlanations (SHAP) to assess driver importance, revealing th at phenology and radiation had the greatest influence on prediction accuracy, fo llowed by temperature and soil moisture.”
GuangdongPeople’s Republic of ChinaA siaCyborgsEmerging TechnologiesMachine Learning
NETL
NSTL
