A hybrid framework: singular value decomposition and kernel ridge regression optimized using mathematical-based fine-tuning for enhancing river water level forecasting
Bibliographic record
Abstract
The precise monitoring and timely alerting of river water levels represent critical measures aimed at safeguarding the well-being and assets of residents in river basins. Achieving this objective necessitates the development of highly accurate river water level forecasts. Hence, a novel hybrid model is provided, incorporating singular value decomposition (SVD) in conjunction with kernel-based ridge regression (SKRidge), multivariate variational mode decomposition (MVMD), and the light gradient boosting machine (LGBM) as a feature selection method, along with the Runge-Kutta optimization (RUN) algorithm for parameter optimization. The L-SKRidge model combines the advantages of both the SKRidge and ridge regression techniques, resulting in a more robust and accurate forecasting tool. By incorporating the linear relationship and regularization techniques of ridge regression with the flexibility and adaptability of the SKRidge algorithm, the L-SKRidge model is able to capture complex patterns in the data while also preventing overfitting. The L-SKRidge method is applied to forecast water levels in the Brook and Dunk Rivers in Canada for two distinct time horizons, specifically one- and three days ahead. Statistical criteria and data visualization tools indicates that the L-SKRidge model has superior efficiency in both the Brook (achieving R = 0.970 and RMSE = 0.051) and Dunk (with R = 0.958 and RMSE = 0.039) Rivers, surpassing the performance of other hybrid and standalone frameworks. The results show that the L-SKRidge method has an acceptable ability to provide accurate water level predictions. This capability can be of significant use to academics and policymakers as they develop innovative approaches for hydraulic control and advance sustainable water resource management.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".