Optimized Photometric Redshifts for the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS)
Bibliographic record
Abstract
Abstract We present the first comprehensive release of photometric redshifts (photo- z's) from the Cosmic Assembly Near-Infrared Deep Extragalactic Legacy Survey (CANDELS) team. We use statistics based upon the Quantile–Quantile (Q–Q) plot to identify biases and signatures of underestimated or overestimated errors in photo- z probability density functions (PDFs) produced by six groups in the collaboration; correcting for these effects makes the resulting PDFs better match the statistical definition of a PDF. After correcting each group’s PDF, we explore three methods of combining the different groups’ PDFs for a given object into a consensus curve. Two of these methods are based on identifying the minimum f-divergence curve, i.e., the PDF that is closest in aggregate to the other PDFs in a set (analogous to the median of an array of numbers). We demonstrate that these techniques yield improved results using sets of spectroscopic redshifts independent of those used to optimize PDF modifications. The best photo- z PDFs and point estimates are achieved with the minimum f-divergence using the best four PDFs for each object (mFDa4) and the hierarchical Bayesian (HB4) methods, respectively. The HB4 photo- z point estimates produced σ NMAD = 0.0227/0.0189 and ∣Δz/(1 + z)∣ > 0.15 outlier fraction = 0.067/0.019 for spectroscopic and 3D Hubble Space Telescope redshifts, respectively. Finally, we describe the structure and provide guidance for the use of the CANDELS photo- z catalogs, which are available at https://archive.stsci.edu/prepds/candels/ .
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 | 0.009 | 0.020 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".