New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines
Abstract In this work, we develop gradient boosting machines (GBMs) for forecasting the SYM‐H index multiple hours ahead using different combinations of solar wind and interplanetary magnetic field (IMF) parameters, derived parameters, and past SYM‐H values. Using Shapley Additive Explanation values...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-08-01
|
| Series: | Space Weather |
| Online Access: | https://doi.org/10.1029/2021SW002928 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841536521618849792 |
|---|---|
| author | Daniel Iong Yang Chen Gabor Toth Shasha Zou Tuija Pulkkinen Jiaen Ren Enrico Camporeale Tamas Gombosi |
| author_facet | Daniel Iong Yang Chen Gabor Toth Shasha Zou Tuija Pulkkinen Jiaen Ren Enrico Camporeale Tamas Gombosi |
| author_sort | Daniel Iong |
| collection | DOAJ |
| description | Abstract In this work, we develop gradient boosting machines (GBMs) for forecasting the SYM‐H index multiple hours ahead using different combinations of solar wind and interplanetary magnetic field (IMF) parameters, derived parameters, and past SYM‐H values. Using Shapley Additive Explanation values to quantify the contributions from each input to predictions of the SYM‐H index from GBMs, we show that our predictions are consistent with physical understanding while also providing insight into the complex relationship between the solar wind and Earth's ring current. In particular, we found that feature contributions vary depending on the storm phase. We also perform a direct comparison between GBMs and neural networks presented in prior publications for forecasting the SYM‐H index by training, validating, and testing them on the same data. We find that the GBMs yield a statistically significant improvement in root mean squared error over the best published black‐box neural network schemes and the Burton equation. |
| format | Article |
| id | doaj-art-d8b5cbbb240d447cbb6a0c1ef1f571f4 |
| institution | Kabale University |
| issn | 1542-7390 |
| language | English |
| publishDate | 2022-08-01 |
| publisher | Wiley |
| record_format | Article |
| series | Space Weather |
| spelling | doaj-art-d8b5cbbb240d447cbb6a0c1ef1f571f42025-01-14T16:27:08ZengWileySpace Weather1542-73902022-08-01208n/an/a10.1029/2021SW002928New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting MachinesDaniel Iong0Yang Chen1Gabor Toth2Shasha Zou3Tuija Pulkkinen4Jiaen Ren5Enrico Camporeale6Tamas Gombosi7Department of Statistics University of Michigan Ann Arbor MI USADepartment of Statistics University of Michigan Ann Arbor MI USADepartment of Climate and Space Sciences and Engineering University of Michigan Ann Arbor MI USADepartment of Climate and Space Sciences and Engineering University of Michigan Ann Arbor MI USADepartment of Climate and Space Sciences and Engineering University of Michigan Ann Arbor MI USADepartment of Climate and Space Sciences and Engineering University of Michigan Ann Arbor MI USACIRES University of Colorado Boulder CO USADepartment of Climate and Space Sciences and Engineering University of Michigan Ann Arbor MI USAAbstract In this work, we develop gradient boosting machines (GBMs) for forecasting the SYM‐H index multiple hours ahead using different combinations of solar wind and interplanetary magnetic field (IMF) parameters, derived parameters, and past SYM‐H values. Using Shapley Additive Explanation values to quantify the contributions from each input to predictions of the SYM‐H index from GBMs, we show that our predictions are consistent with physical understanding while also providing insight into the complex relationship between the solar wind and Earth's ring current. In particular, we found that feature contributions vary depending on the storm phase. We also perform a direct comparison between GBMs and neural networks presented in prior publications for forecasting the SYM‐H index by training, validating, and testing them on the same data. We find that the GBMs yield a statistically significant improvement in root mean squared error over the best published black‐box neural network schemes and the Burton equation.https://doi.org/10.1029/2021SW002928 |
| spellingShingle | Daniel Iong Yang Chen Gabor Toth Shasha Zou Tuija Pulkkinen Jiaen Ren Enrico Camporeale Tamas Gombosi New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines Space Weather |
| title | New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines |
| title_full | New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines |
| title_fullStr | New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines |
| title_full_unstemmed | New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines |
| title_short | New Findings From Explainable SYM‐H Forecasting Using Gradient Boosting Machines |
| title_sort | new findings from explainable sym h forecasting using gradient boosting machines |
| url | https://doi.org/10.1029/2021SW002928 |
| work_keys_str_mv | AT danieliong newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT yangchen newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT gabortoth newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT shashazou newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT tuijapulkkinen newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT jiaenren newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT enricocamporeale newfindingsfromexplainablesymhforecastingusinggradientboostingmachines AT tamasgombosi newfindingsfromexplainablesymhforecastingusinggradientboostingmachines |