Training for the test: Using multi-objective training to improve ANN ensemble forecasts of household residual chlorine in emergencies
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Ensuring that sufficient free residual chlorine (FRC) persist in drinking water throughout the post-distribution period (collection, transport, and household storage) is critical to keeping drinking water safe in emergencies. Probabilistic models like artificial neural network (ANN) ensemble forecasting systems (EFS) have the potential to reproduce the high variability in post-distribution chlorine decay to generate risk-based chlorination guidance, but training with symmetrical error cost functions like mean squared error leads to poor probabilistic performance. This research proposes multi-objective (MO) training to improve the probabilistic performance of ANN-EFS forecasts of post-distribution FRC. Four MO optimizers were tested with combinations of seven objective functions and evaluated using water quality datasets from five emergency settings. MO training substantially improved probabilistic performance over conventional symmetrical error training. The solution that provided the most consistent improvement used preference-based optimization via backpropagation with the following objectives: similarity of mean, variance, and skew, correlation, recall, and precision. This approach achieved high performance at all sites and outperformed all baseline comparisons. These improved models will help humanitarian responders set informed chlorination targets that ensure water remains safe up to the point-of-consumption. This research highlights the importance of tailoring training approaches in ANN drinking water applications and hydroinformatics more broadly.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it