首页|Research from Northeast Institute of Geography and Agroecology Yields New Data o n Machine Learning (Remote estimates of suspended particulate matter in global l akes using machine learning models)
Research from Northeast Institute of Geography and Agroecology Yields New Data o n Machine Learning (Remote estimates of suspended particulate matter in global l akes using machine learning models)
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-A new study on artificial intelligence is now available. According to news originating from Jilin, People's Republic o f China, by NewsRx correspondents, research stated, "Suspended particulate matte r (SPM) in lakes exerts strong impact on light propagation, aquatic ecosystem pr oductivity, which co-varies with nutrients, heavy metal and micro-pollutant in w aters. In lakes, SPM exerts strong absorption and backscattering, ultimately aff ects water leaving signals that can be detected by satellite sensors." The news correspondents obtained a quote from the research from Northeast Instit ute of Geography and Agroecology: "Simple regression models based on specific ba nd or hand ratios have been widely used for SPM estimate in the past with modera te accuracy. There are still rooms for model accuracy improvements, and machine learning models may solve the non-linear relationships between spectral variable and SPM in waters. We assembled more than 16,400 in situ measured SPM in lakes from six continents (excluding the Antarctica continent), of which 9640 samples were matched with Landsat overpasses within ±7 days. Seven machine learning algo rithms and two simple regression methods (linear and partial least squares model s) were used to estimate SPM in lakes and the performance were compared. To over come the problem of imbalance datasets in regression, a Synthetic Minority Over- Sampling technique for regression with Gaussian Noise (SMOGN) was adopted in thi s study. Through comparison, we found that gradient boosting decision tree (GBDT ), random forest (RF), and extreme gradient boosting (XGBoost) models demonstrat ed good spatiotemporal transferability with SMOGN processed dataset, and has pot ential to map SPM at different year with good quality of Landsat land surface re flectance images. In all the tested modeling approaches, the GBDT model has accu rate calibration (n = 6428, R2 = 0.95, MAPE = 29.8 %) from SPM colle cted in 2235 lakes across the world, and the validation (n = 3214, R2 = 0.84, MA PE = 38.8%) also exhibited stable performance. Further, the good pe rformances were also exhibited by RF model with calibration (R2 = 0.93) and vali dation (R2 = 0.86, MAPE = 24.2%) datasets."
Northeast Institute of Geography and Agr oecologyJilinPeople's Republic of ChinaAsiaCyborgsEmerging Technologie sMachine Learning
NETL
NSTL
