Parallel processing and machine learning for long-timescale ambient noise measurements—Illustration with data from the Neptune Ocean Observatory offshore British Columbia
Bibliographic record
Abstract
Seafloor observatories are a rapidly maturing technological approach, enabling the monitoring of marine soundscapes over large areas (basin scale) and long timescales (10 + years). However, the time needed to process broadband measurements, especially over large periods, often acts as a bottleneck. This is particularly true when combining multi-resolution analyses with assessing the impacts of relatively short transients. We are using parallel processing to enable machine learning approaches. To accelerate the computation of spectrograms, we have implemented a parallel processing method that uses the FFT algorithm FFTW3 (http://www.fftw.org/fftw3.pdf), using MPI/C + + on the High Performance Computing facilities at the University of Bath, and compared with spectrogram calculations from well-established software PAMGuide (Merchant et al., Meth. Ecol. Evol. 2015), with Matlab’s Parallel Computing Toolbox. This approach was tested on 1 month of broadband (96 kHz) measurements from the NEPTUNE node at Folger Deep. One month of data can be processed in <3 hours, to a dB accuracy even on short time segments, and that performance increases with the number of parallel processing units. Stability of the parallel approach has been tested with synthetic signals (e.g., chirps) and increasing signal-to-noise ratios. This enables much faster monitoring of long-term trends of important sound metrics.
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 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".