Hyperspectral Estimation of Soil Copper Concentration Based on Improved TabNet Model in the Eastern Junggar Coalfield

IEEE Transactions on Geoscience and Remote Sensing, Journal Year: 2022, Volume and Issue: 60, P. 1 - 20

Published: Jan. 1, 2022

China is the largest coal consumer in world. The massive exploitation and utilization of resources has resulted serious problems heavy metal pollution environmental contamination, such as soil degradation, water pollution, crop damage, even threatening human lives. Therefore, monitoring quickly real time an urgent task at present. This research not only formulated a new preprocessing method enlightened by few-shot learning for hyperspectral data, but also combined it with other soil-related auxiliary information to extract effective from hyperspectrum, end which different regression methods were adopted predict contamination. test used 168 actual samples Eastern Junggar coalfield Xinjiang verification. Since copper trace element corresponding spectral characteristics are affected impurities, improper use may introduce interference or delete useful information, makes model effect unsatisfied. To effectively address above problems, this experiment second-order differential derivation, data enhancement together addition allow more features be entered into model. Next, Attentive Interpretable Tabular Learning (TabNet) was improved three ways using original TabNet models create models. One had best effect, list top 30 according degree importance. Meanwhile, prediction Cu content four convolutional neural networks (CNN) revealed that residual block strongest slightly outperformed model, lacked interpretation input data. Besides, employed pre-processing on various models, found traditional performed (e.g., PLSR) underperformed deep selected optimal compared partial least square (PLSR), network results indicated both CNN better performance approach proposed paper, yielding coefficient determination (R²), root mean error (RMSE) ratio interquartile range (RPIQ) 0.94, 1.341 4.474, respectively. 0.942, 1.324 4.531 dataset.

Language: Английский

---

Prediction of soil heavy metal contents in urban residential areas and the strength of deep learning: A case study of Beijing

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 950, P. 175133 - 175133

Published: July 30, 2024

Language: Английский

---

Predictive modeling of BOD throughout wastewater treatment: a generalizable machine learning approach for improved effluent quality

Environmental Science Water Research & Technology, Journal Year: 2024, Volume and Issue: 10(10), P. 2577 - 2588

Published: Jan. 1, 2024

A machine learning model using easily measured water parameters effectively predicts biochemical oxygen demand across wastewater treatment plants, assisting rapid monitoring and improved effluent quality management.

Language: Английский

---

Retrieval of chromium and mercury concentrations in agricultural soils: Using spectral information, environmental covariates, or a fusion of both?

Ecological Indicators, Journal Year: 2024, Volume and Issue: 167, P. 112594 - 112594

Published: Sept. 14, 2024

Language: Английский

---

Maximizing Heavy Metal Removal and Precious Metal Recovery with Innovative Biowaste-Derived Biosorbents and Biochar

Process Safety and Environmental Protection, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 1, 2024

Language: Английский

---