Author-submitted data information
| ID | 704 |
| Title | Supporting data and code for recovering the primary geochemistry of imperfect zircons using machine learning |
| Creator | Pengfei Lv |
| Subject | Machine Learning, Geochemistry, Zircon, Solid Earth |
| Publisher | Xiukuan Zhao |
| Description | Data set S1 Data original. (separate file) Zircons from granite with full REE, complied from Georoc Database (DIGIS TEAM, 2024). Data set S2 Data processing. (separate file) Input data for machine learning training, obtained by preprocessing Data set S1. Data set S3 updated_data_RF. (separate file) Data set S1 data filled by the trained RF model. Data set S4 updated_data_SVR. (separate file) Data set S1 data filled by the trained SVR model. Data set S5 updated_data_XGB. (separate file) Data set S1 data filled by the trained XGB model. Data set S6 Jack_Hills. (separate file) Jack Hills zircons with full REE, compiled from Bell et al. (2016） Data set S7 updated_Jack_Hills. (separate file) Data set S6 data filled by the trained random forest model. Data set S8 Global_detrital_zircon. (separate file) Detrital zircons with full REE, compiled from Balica et al. (2020). Data set S9 updated_data_Global_detrital_zircon. (separate file) Data set S8 data filled by the trained random forest model. GEOROC zircon geochemistry dataset_1_74837_original. (separate file) GEOROC zircon geochemistry dataset raw dataset. GEOROC zircon geochemistry dataset_1_74837_original_Classified. (separate file) GEOROC zircon geochemistry dataset after tagging. GEOROC zircon geochemistry dataset_1_74837_original_processed. (separate file) Processed GEOROC zircon geochemistry dataset. testdata_recovered_EvalutionValue. (separate file) GEOROC zircon geochemistry dataset used for testing. Code: TrainModelNew This script performs chained imputation of missing values and trains a multi-output random-forest regression model for zircon rare-earth element (REE) data, producing two ready-to-use .pkl model files from the cleaned dataset. Processed_all The code performs quality classification, normalization, missing-value imputation, and outlier correction on zircon rare-earth element (REE) data in Newdata.xlsx. |
| Contributor | Xinyu Zou |
| Date | 13 March, 2025 |
| Type | |
| Format | .xlsx, .py |
| URL | http://www.geophys.ac.cn/ArticleData/20250627ImperfectZircons.zip |
| DOI | 10.12197/2025GA005 |
| Source | |
| Language | eng |
| Relation | |
| Coverage | |
| Rights | Institute of Geology and Geophysics, Chinese Academy of Sciences |