More Efficient and Reliable: Identifying RRab Stars with Blazhko Effect by Deep Convolutional Neural Network
Bibliographic record
Abstract
The physical origin of the Blazhko effect (BL), a phenomenon of a single or multiple periodic modulation(s) of the light curve, is under debate. Efficiently identifying and characterizing the BL is essential in understanding its origins and accounting for its effect on numerous applications of RRabs in the era of large time-domain surveys. In this study, we make use of Resnet 34, a well-known convolutional neural network (CNN) architecture, to identify RRab stars with BL from phased light curves collected from OGLE. Using reliably classified RRabs from frequency analysis to train, validate, and test our model, we show that our CNN method reaches accuracies up to 94%. We then applied our CNN method to some additional RRabs located in the Magellanic Cloud (MC) and the Galactic Bulge (GB), leading to the discovery of 113 and 2496 BL candidates, respectively. The identification accuracy for the MC Sample is estimated to be 91% after cross-matching the CNN classification results with those from frequency analysis. Similarly, the light-curve parameters of these classified BL/non-BL candidates by our CNN method from the GB region resemble those observed in the literature, confirming the reliability of our CNN classifications. Our CNN method is subject to issues related to light-curve quality and sampling, but its overall reliance on light-curve quality is comparable to that of frequency analysis. Furthermore, we find that BL modulation could be primarily characterized by variations in light-curve structure.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| 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 | 0.000 |
| 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".