Pith. sign in

REVIEW 4 major objections 7 minor 112 references

TDEs can be found from light curves alone if classifiers treat features as uncertain distributions rather than fixed numbers.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-31 05:04 UTC pith:27AQCCYG

load-bearing objection Solid ZTF methods paper: PRF vs XGBoost on host-agnostic TDE features is carefully done; the “essential for Rubin” claim rests too heavily on one synthetic degradation test. the 4 major comments →

arxiv 2607.28510 v1 pith:27AQCCYG submitted 2026-07-30 astro-ph.HE astro-ph.IM

Uncertainty-Aware Tidal Disruption Event Classification : A Host-Agnostic Probabilistic Random Forest Approach

Vysakh Anilkumar , Sjoert van Velzen , Marek Kowalski , Simeon Reusch This is my paper
classification astro-ph.HE astro-ph.IM
keywords tidal disruption eventsphotometric classificationprobabilistic random foresthost-agnosticmeasurement uncertaintyZTFRubin Observatorynuclear transients
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Tidal disruption events are rare flares that mark a star being torn apart by a supermassive black hole. Upcoming surveys will find far more of them than can ever receive spectra, and many will be too faint for host-galaxy tricks that current classifiers rely on. This paper shows that eleven photometric light-curve features—rise and decay times, temperature evolution, and pre-flare variability—are enough to separate TDEs from supernovae and AGN without any host information. The key is a Probabilistic Random Forest that treats every feature as a distribution rather than a single number, so measurement noise is carried into the decision. Against a standard deterministic booster the probabilistic model is more stable on ambiguous cases and rejects substantially more false positives when completeness is prioritized. Applied to archival ZTF data it recovers eleven new candidates and flags three previously mislabeled sources. The result is a host-independent pipeline whose conservative behavior is designed for the low-signal regime that will dominate the next decade of transient astronomy.

Core claim

TDEs can be reliably identified from photometric light-curve features alone in a host-agnostic framework. A Probabilistic Random Forest that ingests feature uncertainties as distributions yields higher stability for ambiguous candidates and rejects far more false positives in the high-recall regime than a deterministic XGBoost baseline, establishing that uncertainty-aware classifiers are required to avoid brittle, overconfident errors at low signal-to-noise.

What carries the argument

The Probabilistic Random Forest (PRF): a bagging ensemble that routes each object through every tree split according to the cumulative distribution of its measured features, so measurement variance is propagated into both the predicted class probabilities and the feature-level contributions.

Load-bearing premise

That aggressively oversampling the handful of real TDEs by linear interpolation in feature-and-uncertainty space produces synthetic examples realistic enough for the learned boundaries to hold on genuinely faint, sparsely sampled light curves.

What would settle it

Apply the identical PRF and XGBoost pipelines, without re-tuning, to a large set of simulated Rubin-like nuclear light curves whose true labels are known; if the PRF no longer shows higher stability and lower false-positive rate at faint magnitudes, the central claim fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Host-galaxy cuts and WISE AGN indicators become optional rather than required for photometric TDE selection.
  • In high-completeness searches the PRF can cut roughly a third more contaminants than a deterministic booster at the same recall.
  • Eleven new archival candidates and three reclassified labeled sources become concrete spectroscopic follow-up targets.
  • Faint-end classifiers for the Rubin alert stream should propagate feature uncertainties rather than treat medians as exact.
  • PRF and XGBoost occupy complementary regimes and can be combined as an orthogonal filter.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same uncertainty-propagation idea should transfer to other rare nuclear classes (changing-look AGN, partial TDEs) whose light curves are equally sparse.
  • Once physically motivated light-curve injections replace linear SMOTE, the same eleven features could support early-time rather than full-light-curve classification.
  • A gradient-boosting variant that natively accepts feature distributions would combine XGBoost’s recall with PRF’s noise handling.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 7 minor

Summary. The paper develops a host-agnostic photometric classifier for tidal disruption events (TDEs) in the ZTF nuclear-transient stream. Eleven light-curve features (rise/decay timescales, blackbody temperature evolution, pre-transient and plateau fractional rms, plateau strength/SNR, etc.) are extracted via a multi-band MCMC fit that extends the van Velzen et al. (2021) model with a late-time plateau. Classification is performed with a Probabilistic Random Forest (PRF) that treats features as distributions, benchmarked against deterministic XGBoost under a repeated stratified Leave-One-Out Cross-Validation (K=59 TDEs, 250 seeds) with uncertainty-aware SMOTE. The authors report complementary regimes (XGBoost higher recall at high precision; PRF fewer false positives at high recall and more stable probabilities for ambiguous sources), recover 10/12 post-2023 spectroscopic TDEs, and identify 11 new archival candidates plus 3 possible label contaminants. A single-source synthetic flux-degradation test is used to argue that PRF degrades more gracefully than XGBoost and is therefore essential for the low-S/N Rubin regime.

Significance. Host-agnostic, purely photometric TDE selection is a genuine and timely need for LSST/Rubin, where host WISE/AGN indicators and spectroscopic follow-up will be unavailable for most faint nuclear flares. The work’s concrete strengths include: (i) explicit propagation of feature posteriors into a PRF rather than point estimates; (ii) a carefully documented multi-class LOOCV with 250 repetitions, threshold scenarios, confusion matrices, and SHAP importances; (iii) an independent post-cutoff TDE test set; (iv) a controlled signal-degradation experiment; and (v) public release of forced-photometry light curves and per-source class probabilities on Zenodo. If the stability and false-positive advantages of uncertainty-aware classification hold under more realistic Rubin-like sampling, the methodological message is important for the broader transient-classification community. The archival candidate list is a useful byproduct even if purity remains unquantified on the unlabeled pool.

major comments (4)
  1. [§6.2, Fig. 10; Abstract; §7] §6.2 and Fig. 10 are the primary quantitative support for the abstract’s and §7’s claim that uncertainty-aware PRF classifiers are “essential for the Rubin era.” The experiment uses a single high-confidence template (TDE2021axu), scales its GP mean flux, and re-injects noise drawn from that same ZTF light curve’s empirical flux–error relation. Cadence, baseline coverage, and morphological diversity therefore remain pure ZTF. A single-template, same-survey noise model cannot establish that the graceful-vs-abrupt behaviour generalises across the TDE population, under Rubin-like sparse sampling, or under the SMOTE-trained boundaries learned on bright ZTF features. Either expand the degradation test to a multi-source ensemble and/or cadence-mismatched noise, or substantially qualify the Rubin-forward claim so that it is presented as a motivated hypothesis rather than a demonstrated necessity
  2. [§4.3.1] §4.3.1 applies aggressive uncertainty-aware SMOTE that oversamples the 59 real TDEs (and SNe) up to the AGN count inside every LOOCV training fold. The paper itself notes that linear interpolation in feature-and-uncertainty space “is not ideal” and defers physically motivated injections to future work. Because the reported decision boundaries, SHAP rankings, and threshold metrics are learned on this heavily synthetic minority population, the generalisation claim to real faint/low-S/N events rests on an untested assumption. At minimum, the manuscript should quantify sensitivity to the oversampling factor (e.g., metrics vs. milder SMOTE ratios or no SMOTE with class weights) and state clearly which performance numbers are most affected. Without that, the LOOCV metrics cannot be read as reliable estimates of behaviour on real Rubin-like light curves.
  3. [§4.1–4.4; §6.1] The PRF–XGBoost comparison confounds two distinct design choices: (i) native uncertainty propagation vs. point estimates, and (ii) bagging vs. sequential gradient boosting. §4.4 and §6.1 attribute stability and false-positive rejection largely to uncertainty handling, yet the bagging architecture alone (independent trees, averaged predictions) already reduces variance under LOOCV perturbations. A standard (non-probabilistic) Random Forest baseline, or an ablation that feeds PRF median features without uncertainties, is needed to isolate how much of the reported advantage is truly due to treating features as distributions. Without that control, the central methodological claim—that uncertainty awareness is what makes the classifier robust—is only partially supported.
  4. [§4.3.3; §5.1–5.2; Table 3] §5.1 explicitly states that reported probabilities are uncalibrated classifier outputs and “should not be interpreted as posterior probabilities,” yet the high-precision / balanced / high-recall scenarios (§4.3.3, Table 3, Fig. 4–5) and the archival search (§5.2) apply absolute probability thresholds (e.g. ≈0.85) as if they were meaningful on an absolute scale. Thresholds chosen on uncalibrated scores are fold- and architecture-dependent; the complementary-regime narrative and the candidate cuts therefore need either a calibration step (Platt/isotonic on held-out folds) or a clear statement that all thresholds are relative operating points only, with purity/completeness re-estimated under that caveat when the ensemble is applied to unlabeled data.
minor comments (7)
  1. [Table 3] Table 3 reports asymmetric uncertainties with occasional identical upper/lower signs (e.g. PRF high-recall Recall 0.966+0.017+0.000). Clarify whether these are 16th/84th percentiles and correct any typographical sign errors.
  2. [Fig. 3] Figure 3 violin plots are dense; labeling only a subset of TDE IDs or providing an interactive/supplementary version would improve readability. Cross-reference to Table A1 is helpful but easy to miss.
  3. [§3.3, Eq. (5)] Eq. (5) normalises excess variance by F_peak rather than by mean host flux; a one-sentence justification relative to the more common f_var definition would help non-specialist readers.
  4. [§3.1] The plateau sigmoid timescale is fixed at τ_plat = 3.264 days (§3.1) to reach 99% of F_plat near t_peak+30 d. State whether results are sensitive to this choice.
  5. [References; Table 4–5 notes] Several in-text citations and TNS report formatting are inconsistent (e.g. mixed “TNS Classification Report” styles). A pass for uniform reference formatting would help.
  6. [§8] Data availability (§8) is commendable. Adding a short README pointer to which catalog columns map to the 11 features and the PRF/XGB probability columns would lower the barrier for reuse.
  7. [§2.2] In §2.2, the zero-point cut is written as |log10 ZP / med(ZP)| < 0.4; confirm the exact expression and units so the cut is reproducible.

Circularity Check

1 steps flagged

Empirical ML comparison with external labels; only mild coauthor method reuse, no derivation-by-construction.

specific steps
  1. self citation load bearing [§2.1; §3.1 (Eq. 1–3)]
    "This filter, originally used by van Velzen et al. (2019a, 2021); Hammerstein et al. (2023b)... Our base model for the transient component is adapted from the light curve model presented in van Velzen et al. (2021), which is characterized by a Gaussian rise to the peak followed by an exponential decay. We further extend this model to account for the persistent, late-time flux plateau..."

    The nuclear selection filter and the parametric light-curve model are taken from prior papers with overlapping authors (van Velzen, Hammerstein, Reusch). This is ordinary method reuse, not a uniqueness import that forces the PRF-vs-XGBoost result; labels and LOOCV metrics remain external. Flagged only as minor coauthor dependence on inputs, not as a circular derivation of the main claim.

full rationale

The paper’s load-bearing claims are empirical classifier comparisons (PRF vs XGBoost under LOOCV) and an archival candidate search, not first-principles predictions. Training labels are taken from external spectroscopic catalogs (BTS, TNS, manyTDE, WISE AGN, milliquas). Features are fitted per light curve via MCMC; the classifiers are then scored on held-out LOOCV folds or on post-cutoff TDEs never used in training. Nothing algebraically forces a ‘prediction’ from a fitted constant. Mild self-citation exists for the nuclear-alert filter and the van Velzen et al. (2021) light-curve ansatz (coauthor overlap), but those are standard methodological inputs, not uniqueness theorems or self-definitional closures. The Rubin-era degradation argument (Fig. 10) is a controlled experiment on one template—not circular, though limited in scope. Overall circularity is negligible.

Axiom & Free-Parameter Ledger

6 free parameters · 5 axioms · 0 invented entities

The central claim rests on standard transient-astronomy assumptions (nuclear offset cut, blackbody SED, spectroscopic labels as ground truth), on the Reis et al. PRF formalism, and on several analysis choices (SMOTE multiplicity, probability thresholds, freeze-out of poorly constrained plateau parameters) that are not derived from first principles. No new physical entities are postulated.

free parameters (6)
  • PRF keep_proba threshold = 0.05
    Branch-pruning probability set to 0.05 following Reis et al.; controls how uncertainty is truncated.
  • PRF max_depth / n_estimators = depth=8, n=100
    Fixed at 8 and 100 to control overfitting on the small TDE sample.
  • SMOTE oversampling target = match AGN count
    TDE and SN classes are synthetically expanded to equal the AGN count (~5113), an aggressive choice acknowledged by the authors.
  • High-precision / balanced / high-recall probability thresholds = scenario-dependent cuts
    Chosen post-hoc to hit ~80% purity, ~85% purity, and ~95% completeness operating points.
  • Plateau sigmoid timescale tau_plat = 3.264 days
    Hard-coded so the sigmoid reaches 99% of F_plat at t_peak+30 d.
  • Nuclear offset cut = <0.5 arcsec
    Sources retained only if <0.5 arcsec from host nucleus; defines the input sample.
axioms (5)
  • domain assumption Spectroscopic labels from BTS/TNS/manyTDE/WISE/milliquas are sufficiently pure ground truth for supervised training.
    Section 2.3; three later re-labeled sources show the labels are imperfect.
  • domain assumption A Gaussian-rise + exponential-decay + late-time sigmoid plateau, scaled by a piecewise-linear blackbody temperature, adequately captures the photometric diversity of TDEs, SNe and AGN flares.
    Equations 1–3, Section 3.1; model failures drop 581+ sources.
  • domain assumption Feature measurement uncertainties can be treated as independent Gaussian (or infinite) distributions that the PRF propagates correctly.
    Section 4.1; core modeling choice of the PRF.
  • ad hoc to paper Linear interpolation in feature-and-uncertainty space (SMOTE) yields synthetic TDEs that do not distort the decision boundary for real low-S/N events.
    Section 4.3.1; authors flag this as non-ideal.
  • domain assumption Host-galaxy information can be entirely omitted without loss of essential discriminative power for the faint Rubin population.
    Stated motivation throughout Introduction and Abstract.

pith-pipeline@v1.2.0-daily-grok45 · 34450 in / 3423 out tokens · 60813 ms · 2026-07-31T05:04:19.032523+00:00 · methodology

0 comments
read the original abstract

The classification of Tidal Disruption Events in large-scale photometric surveys is challenging because deterministic machine learning models produce overconfident misclassifications under varying data quality and low signal-to-noise conditions. Existing lightcurve-based approaches fail to incorporate measurement uncertainties, consequently generating brittle outputs. We present a host-agnostic, uncertainty-aware classification framework using a Probabilistic Random Forest (PRF). Our pipeline extracts 11 characteristic features from nuclear transients in the ZTF alert stream, including rise and decay timescales, blackbody temperature evolution, and pre-transient variability metrics. Relying exclusively on photometric data without host galaxy information ensures effectiveness for the faint transient population expected from the Rubin Observatory. The PRF, which treats feature measurements as distributions, was evaluated against XGBoost through Leave-One-Out Cross-Validation. It yields higher stability and robustness for ambiguous candidates than XGBoost. The two classifiers occupy complementary regimes. XGBoost achieves higher recall in balanced and high-precision scenarios, where its rigid decision boundaries efficiently isolate TDE-like sources, while PRF rejects more false positives by penalizing sources with large feature uncertainties. Applying this framework to archival data identified 11 new candidate TDEs from the unclassified population and 3 potential photometric TDEs within existing training labels previously misclassified as supernovae or active galactic nuclei. This work demonstrates that TDEs can be reliably identified using photometric lightcurve features, providing a host-independent framework. Uncertainty-aware, probabilistic classifiers are essential for the Rubin era to prevent the overconfident misclassifications inherent in deterministic models at low signal-to-noise.

Figures

Figures reproduced from arXiv: 2607.28510 by Marek Kowalski, Simeon Reusch, Sjoert van Velzen, Vysakh Anilkumar.

Figure 1
Figure 1. Figure 1: The top panel shows the multi-band lightcurve model fitted to 𝑟 and 𝑔 bands of the source TDE2021nwa. The black dot-dash line line shows the component that models the plateau of the TDE. The bottom panel shows the estimated evolution of temperature using the model in equation 3. log10 𝑇 (𝑡) =    log10 𝑇peak 𝑡 < 𝑡peak log10 𝑇peak + 𝑑 log10 𝑇 𝑑𝑡 (𝑡 − 𝑡peak) 𝑡peak ≤ 𝑡 ≤ 𝑡plat log10 𝑇plat 𝑡 ≥ 𝑡plat (3… view at source ↗
Figure 2
Figure 2. Figure 2: The top panel demonstrates PRF constructing independent trees simultaneously, using probabilistic propagation to route data points across multiple branches using measurement uncertainties at each nodes. Crossed-out paths indicate branches where the propagation probability falls below the defined prediction threshold. The lower panel illustrates XGBoost constructing trees sequentially, relying on determinis… view at source ↗
Figure 3
Figure 3. Figure 3: This figure shows the results of the LOOCV using violin plots for PRF classifier (left panel) and XGBoost classifier (right panel). The y-axis for both plots display the predicted probability by the corresponding ML algorithm. The y-axis are the TDE IDs assigned to each of the 59 TDEs used in this work. The corresponding ZTF IDs and the TNS names are show in Table A1. MNRAS 000, 1–20 (2026) [PITH_FULL_IMA… view at source ↗
Figure 4
Figure 4. Figure 4: This figure displays the threshold tuning results for the PRF (left panel) and XGBoost (right panel) classifiers. The plotted lines represent the median Precision, Recall, and F1 Score as a function of the probability threshold across the 250 LOOCV iterations. The shaded regions denote the 16th and 84th percentile uncertainty intervals for each evaluated metric. non-TDEs that the XGBoost model overfits as … view at source ↗
Figure 5
Figure 5. Figure 5: Confusion matrix plots for different threshold selection scenarios for both classifiers. For each blocks in the confusion matrix, the numbers in the bracket shows the counts of the AGN and SNe belonging to that particular block (There are 59 TDEs, 1029 SNE and 4357 AGN). The colors in the confusion matrix are normalized column-wise with the normalization scale shown at the bottom. Top panel shows the confu… view at source ↗
Figure 6
Figure 6. Figure 6: This figure shows the global feature importance for both classifiers. The features are ranked based on their global mean absolute SHAP values, which represent the average contribution of each feature to the model’s predictions across the dataset. Each bar is divided into three colored segments corresponding to the three classes: TDE (blue), SN (orange), and AGN (red). The length of each segment indicates t… view at source ↗
Figure 7
Figure 7. Figure 7: This figure compares the properties of the known TDE training sample to those of the newly identified candidates. The left panel displays the inferred physical parameter distributions (𝑡rise, 𝑇peak, 𝑡decay), demonstrating the morphological overlap between the two populations. The right panel presents the stacked distribution of peak AB magnitudes for both the known TDEs and the new candidates. The new cand… view at source ↗
Figure 8
Figure 8. Figure 8: A detailed view of the TDE probability distributions originally presented in [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
Figure 10
Figure 10. Figure 10: Classification robustness of the models under simulated signal￾to-noise degradation. The x-axis represents the systematically reduced peak flux of the reference transient TDE2021axu, mapped to apparent AB mag￾nitude. The markers indicate the median predicted TDE probability across 100 randomized photometric noise iterations at each magnitude bin, and the shaded regions represent the corresponding IQR for … view at source ↗
Figure 9
Figure 9. Figure 9: (a) ZTF lightcurve of TDE2022emf. Thicker lines represent the median estimates of the fit, while thinner lines represent random draws from the posterior distribution generated via MCMC sampling. (b) The top panel displays the SHAP waterfall plot for the XGBoost classifier, and the bottom panel displays the same for the PRF classifier. Darker shades indicate a posi￾tive feature contribution, while lighter s… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

112 extracted references · 2 canonical work pages · 1 internal anchor

  1. [1]

    , keywords =

    Probabilistic Random Forest: A Machine Learning Algorithm for Noisy Data Sets. , keywords =. doi:10.3847/1538-3881/aaf101 , archivePrefix =. 1811.05994 , primaryClass =

    Pith/arXiv arXiv

  2. [2]

    , keywords =

    Seventeen Tidal Disruption Events from the First Half of ZTF Survey Observations: Entering a New Era of Population Studies. , keywords =. doi:10.3847/1538-4357/abc258 , archivePrefix =. 2001.01409 , primaryClass =

    Pith/arXiv arXiv 2001

  3. [3]

    , keywords =

    The First Tidal Disruption Flare in ZTF: From Photometric Selection to Multi-wavelength Characterization. , keywords =. doi:10.3847/1538-4357/aafe0c , archivePrefix =. 1809.02608 , primaryClass =

    Pith/arXiv arXiv

  4. [4]

    , keywords =

    Integral Field Spectroscopy of 13 Tidal Disruption Event Hosts from the Zwicky Transient Facility Survey. , keywords =. doi:10.3847/1538-4357/acfb84 , archivePrefix =. 2307.15705 , primaryClass =

    Pith/arXiv arXiv

  5. [5]

    , keywords =

    tdescore: An Accurate Photometric Classifier for Tidal Disruption Events. , keywords =. doi:10.3847/2041-8213/ad3337 , archivePrefix =. 2312.00139 , primaryClass =

    Pith/arXiv arXiv 2041

  6. [6]

    , keywords =

    Transient processing and analysis using AMPEL: alert management, photometry, and evaluation of light curves. , keywords =. doi:10.1051/0004-6361/201935634 , archivePrefix =. 1904.05922 , primaryClass =

    Pith/arXiv arXiv 1904

  7. [7]

    2024 , address =

    Reusch, Simeon , title =. 2024 , address =

    2024

  8. [8]

    arXiv e-prints , keywords =

    A New Forced Photometry Service for the Zwicky Transient Facility. arXiv e-prints , keywords =. doi:10.48550/arXiv.2305.16279 , archivePrefix =. 2305.16279 , primaryClass =

    Pith/arXiv arXiv doi:10.48550/arxiv.2305.16279

  9. [9]

    , keywords =

    Measuring Reddening with Sloan Digital Sky Survey Stellar Spectra and Recalibrating SFD. , keywords =. doi:10.1088/0004-637X/737/2/103 , archivePrefix =. 1012.4804 , primaryClass =

    Pith/arXiv arXiv

  10. [10]

    , keywords =

    Correcting for the Effects of Interstellar Extinction. , keywords =. doi:10.1086/316293 , archivePrefix =. astro-ph/9809387 , primaryClass =

    Pith/arXiv arXiv

  11. [11]

    , keywords =

    emcee: The MCMC Hammer. , keywords =. doi:10.1086/670067 , archivePrefix =. 1202.3665 , primaryClass =

    Pith/arXiv arXiv

  12. [12]

    UVIT catalogue of open clusters with machine learning-based membership using Gaia EDR3 astrometry

    UOCS - III. UVIT catalogue of open clusters with machine learning-based membership using Gaia EDR3 astrometry. , keywords =. doi:10.1093/mnras/stab213 , archivePrefix =. 2101.07122 , primaryClass =

    Pith/arXiv arXiv

  13. [13]

    , keywords =

    The probabilistic random forest applied to the selection of quasar candidates in the QUBRICS survey. , keywords =. doi:10.1093/mnras/stab1867 , archivePrefix =. 2106.12990 , primaryClass =

    Pith/arXiv arXiv

  14. [14]

    , keywords =

    A new automated tool for the spectral classification of OB stars. , keywords =. doi:10.1051/0004-6361/202040224 , archivePrefix =. 2110.10669 , primaryClass =

    Pith/arXiv arXiv

  15. [15]

    , keywords =

    Massive young stellar objects in the Local Group irregular galaxy NGC 6822 identified using machine learning. , keywords =. doi:10.1093/mnras/stab2386 , archivePrefix =. 2108.07105 , primaryClass =

    Pith/arXiv arXiv

  16. [16]

    Classifying extragalactic sources using a probabilistic random forest supervised machine learning algorithm

    The VMC Survey - LI. Classifying extragalactic sources using a probabilistic random forest supervised machine learning algorithm. , keywords =. doi:10.1093/mnras/staf080 , archivePrefix =. 2501.08196 , primaryClass =

    Pith/arXiv arXiv

  17. [17]

    , keywords =

    Dusty Stellar Source Classification by Implementing Machine Learning Methods Based on Spectroscopic Observations in the Magellanic Clouds. , keywords =. doi:10.3847/1538-4357/adceeb , archivePrefix =. 2504.14332 , primaryClass =

    Pith/arXiv arXiv

  18. [18]

    , keywords =

    A Comprehensive and Uniform Sample of Broad-line Active Galactic Nuclei from the SDSS DR7. , keywords =. doi:10.3847/1538-4365/ab298b , archivePrefix =. 1906.05597 , primaryClass =

    Pith/arXiv arXiv 1906

  19. [19]

    , keywords =

    The WISE AGN Catalog. , keywords =. doi:10.3847/1538-4365/aaa00a , archivePrefix =. 1706.09901 , primaryClass =

    Pith/arXiv arXiv

  20. [20]

    The Zwicky Transient Facility Bright Transient Survey. I. Spectroscopic Classification and the Redshift Completeness of Local Galaxy Catalogs. , keywords =. doi:10.3847/1538-4357/ab8943 , archivePrefix =. 1910.12973 , primaryClass =

    Pith/arXiv arXiv 1910

  21. [21]

    The Zwicky Transient Facility Bright Transient Survey. II. A Public Statistical Sample for Exploring Supernova Demographics. , keywords =. doi:10.3847/1538-4357/abbd98 , archivePrefix =. 2009.01242 , primaryClass =

    Pith/arXiv arXiv 2009

  22. [22]

    , keywords =

    Fundamental scaling relationships revealed in the optical light curves of tidal disruption events. , keywords =. doi:10.1093/mnras/stad3001 , archivePrefix =. 2308.08255 , primaryClass =

    Pith/arXiv arXiv

  23. [23]

    arXiv e-prints , keywords =

    SMOTE: Synthetic Minority Over-sampling Technique. arXiv e-prints , keywords =. doi:10.48550/arXiv.1106.1813 , archivePrefix =. 1106.1813 , primaryClass =

    Pith/arXiv arXiv doi:10.48550/arxiv.1106.1813

  24. [24]

    Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages =

    Chen, Tianqi and Guestrin, Carlos , title =. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages =. 2016 , isbn =. doi:10.1145/2939672.2939785 , abstract =

    arXiv 2016

  25. [25]

    The Astronomer's Telegram , author =

    The. The Astronomer's Telegram , author =. 2018 , note =

    2018

  26. [26]

    and Lee, Su-In , title =

    Lundberg, Scott M. and Lee, Su-In , title =. Proceedings of the 31st International Conference on Neural Information Processing Systems , pages =. 2017 , isbn =

    2017

  27. [27]

    The Open Journal of Astrophysics , keywords =

    The Million Quasars (Milliquas) Catalogue, v8. The Open Journal of Astrophysics , keywords =. doi:10.21105/astro.2308.01505 , archivePrefix =. 2308.01505 , primaryClass =

    Pith/arXiv arXiv

  28. [28]

    Nature , author =

    Tidal disruption of stars by black holes of 106-108 solar masses in nearby galaxies , volume =. Nature , author =. 1988 , note =. doi:10.1038/333523a0 , abstract =

    doi:10.1038/333523a0 1988

  29. [29]

    ARA&A , author =

    Tidal. ARA&A , author =. 2021 , note =. doi:10.1146/annurev-astro-111720-030029 , abstract =

    doi:10.1146/annurev-astro-111720-030029 2021

  30. [30]

    and Kulkarni, Shrinivas R

    Bellm, Eric C. and Kulkarni, Shrinivas R. and Graham, Matthew J. and Dekany, Richard and Smith, Roger M. and Riddle, Reed and Masci, Frank J. and Helou, George and Prince, Thomas A. and Adams, Scott M. and Barbarino, C. and Barlow, Tom and Bauer, James and Beck, Ron and Belicki, Justin and Biswas, Rahul and Blagorodnova, Nadejda and Bodewits, Dennis and B...

  31. [31]

    PASP , author =

    The. PASP , author =. 2020 , note =. doi:10.1088/1538-3873/ab4ca2 , abstract =

    doi:10.1088/1538-3873/ab4ca2 2020

  32. [32]

    ApJ , author =

    The. ApJ , author =. 2020 , note =. doi:10.3847/1538-4357/ab6989 , abstract =

    doi:10.3847/1538-4357/ab6989 2020

  33. [33]

    PASP , author =

    The. PASP , author =. 2019 , note =. doi:10.1088/1538-3873/ab006c , abstract =

    doi:10.1088/1538-3873/ab006c 2019

  34. [34]

    Ashley and Berger, Edo and Gezari, Suvi and van Velzen, Sjoert and Nicholl, Matt and Blanchard, Peter K

    Gomez, Sebastian and Villar, V. Ashley and Berger, Edo and Gezari, Suvi and van Velzen, Sjoert and Nicholl, Matt and Blanchard, Peter K. and Alexander, Kate. D. , month = jun, year =. Identifying. ApJ , publisher =. doi:10.3847/1538-4357/acc535 , abstract =

    doi:10.3847/1538-4357/acc535

  35. [35]

    A&A , author =

    Early. A&A , author =. 2026 , note =. doi:10.1051/0004-6361/202556558 , abstract =

    doi:10.1051/0004-6361/202556558 2026

  36. [36]

    , keywords =

    ALeRCE light curve classifier: Tidal disruption event expansion pack. , keywords =. doi:10.1051/0004-6361/202451951 , archivePrefix =. 2503.19698 , primaryClass =

    Pith/arXiv arXiv

  37. [37]

    2024 , pages =

    , author =. 2024 , pages =. doi:10.1093/mnras/stae1253 , abstract =

    doi:10.1093/mnras/stae1253 2024

  38. [38]

    A&A , author =

    A. A&A , author =. 2025 , note =. doi:10.1051/0004-6361/202556839 , abstract =

    doi:10.1051/0004-6361/202556839 2025

  39. [39]

    2019 , note =

    ApJ , author =. 2019 , note =. doi:10.3847/1538-4357/ab042c , abstract =

    doi:10.3847/1538-4357/ab042c 2019

  40. [40]

    and Sfaradi, Itai and Andreoni, Igor and Charalampopoulos, Panos and Chornock, Ryan and Gezari, Suvi and Mo, Geoffrey and Yao, Yuhan and Anumarlapudi, Akash and Bellm, Eric C

    Stein, Robert and Carney, Jonathan and Ward, Charlotte and Margutti, Raffaella and Hall, Xander J. and Sfaradi, Itai and Andreoni, Igor and Charalampopoulos, Panos and Chornock, Ryan and Gezari, Suvi and Mo, Geoffrey and Yao, Yuhan and Anumarlapudi, Akash and Bellm, Eric C. and Bloom, Joshua S. and Busmann, Malte and Caiazzo, Ilaria and Cenko, S. Bradley ...

    2026

  41. [41]

    , keywords =

    Identifying Transient Hosts in LSST's Deep Drilling Fields with Galaxy Catalogs. , keywords =. doi:10.3847/1538-4357/ae4e1a , archivePrefix =. 2511.22634 , primaryClass =

    arXiv

  42. [42]

    Wang, Mengye and Ma, Yiqiu and Wu, Qingwen and Jiang, Ning , month = jan, year =. An. ApJ , publisher =. doi:10.3847/1538-4357/ad0bfb , abstract =

    doi:10.3847/1538-4357/ad0bfb

  43. [43]

    The optical, UV-plateau and X-ray tidal disruption event luminosity functions reproduced from first principles

    Mummery, Andrew and Velzen, Sjoert van , month = oct, year =. The optical,. doi:10.48550/arXiv.2410.17087 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2410.17087

  44. [44]

    , author =

    The maximum mass of a black hole which can tidally disrupt a star: measuring black hole spins with tidal disruption events , volume =. , author =. 2023 , note =. doi:10.1093/mnras/stad3636 , abstract =

    doi:10.1093/mnras/stad3636 2023

  45. [45]

    and Metzger, Brian D

    van Velzen, Sjoert and Stone, Nicholas C. and Metzger, Brian D. and Gezari, Suvi and Brown, Thomas M. and Fruchter, Andrew S. , month = jun, year =. Late-time. ApJ , publisher =. doi:10.3847/1538-4357/ab1844 , abstract =

    doi:10.3847/1538-4357/ab1844

  46. [46]

    TNS Classification Report , keywords =

    Transient Classification Report for 2024-06-07. TNS Classification Report , keywords =

    2024

  47. [47]

    TNS Classification Report , keywords =

    SGLF Transient Classification Report for 2024-02-23. TNS Classification Report , keywords =

    2024

  48. [48]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2023-11-16. TNS Classification Report , keywords =

    2023

  49. [49]

    TNS Classification Report , keywords =

    Transient Classification Report for 2023-12-21. TNS Classification Report , keywords =

    2023

  50. [50]

    TNS Classification Report , keywords =

    Transient Classification Report for 2023-09-18. TNS Classification Report , keywords =

    2023

  51. [51]

    TNS Classification Report , keywords =

    GOTO Transient Classification Report for 2024-02-11. TNS Classification Report , keywords =

    2024

  52. [52]

    TNS Classification Report , keywords =

    Transient Classification Report for 2023-12-13. TNS Classification Report , keywords =

    2023

  53. [53]

    TNS , keywords =

    ZTF Transient Classification Report for 2024-02-03. TNS , keywords =

    2024

  54. [54]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2023-12-01. TNS Classification Report , keywords =

    2023

  55. [55]

    TNS Classification Report , keywords =

    StarDestroyers Transient Classification Report for 2023-12-30. TNS Classification Report , keywords =

    2023

  56. [56]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2024-05-24. TNS Classification Report , keywords =

    2024

  57. [57]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2023-07-21. TNS Classification Report , keywords =

    2023

  58. [58]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2022-05-09. TNS Discovery Report , keywords =

    2022

  59. [59]

    TNS Discovery Report , keywords =

    ZTF Transient Discovery Report for 2022-05-04. TNS Discovery Report , keywords =

    2022

  60. [60]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2022-11-19. TNS Discovery Report , keywords =

    2022

  61. [61]

    TNS Discovery Report , keywords =

    GOTO Transient Discovery Report for 2021-05-10. TNS Discovery Report , keywords =

    2021

  62. [62]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2021-03-20. TNS Discovery Report , keywords =

    2021

  63. [63]

    TNS Discovery Report , keywords =

    ATLAS Transient Discovery Report for 2021-01-18. TNS Discovery Report , keywords =

    2021

  64. [64]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2021-11-10. TNS Discovery Report , keywords =

    2021

  65. [65]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2020-11-11. TNS Discovery Report , keywords =

    2020

  66. [66]

    TNS Classification Report , keywords =

    C-SNAILS Transient Classification Report for 2019-08-22. TNS Classification Report , keywords =

    2019

  67. [67]

    TNS Discovery Report , keywords =

    ALeRCE/ZTF Transient Discovery Report for 2020-12-15. TNS Discovery Report , keywords =

    2020

  68. [68]

    TNS Discovery Report , keywords =

    ZTF Transient Discovery Report for 2019-06-03. TNS Discovery Report , keywords =

    2019

  69. [69]

    , keywords =

    Estimating photometric redshifts for galaxies from the DESI Legacy Imaging Surveys with Bayesian neural networks trained by DESI EDR. , keywords =. doi:10.1093/mnras/stae2713 , archivePrefix =. 2412.02390 , primaryClass =

    Pith/arXiv arXiv

  70. [70]

    Research Notes of the American Astronomical Society , keywords =

    VLA Observations Confirm AT 2023mfm as an Off-nuclear Tidal Disruption Event. Research Notes of the American Astronomical Society , keywords =. doi:10.3847/2515-5172/ae7a6a , archivePrefix =. 2606.06595 , primaryClass =

    Pith/arXiv arXiv

  71. [71]

    , keywords =

    Spectral Classification and Redshift Measurement for the SDSS-III Baryon Oscillation Spectroscopic Survey. , keywords =. doi:10.1088/0004-6256/144/5/144 , archivePrefix =. 1207.7326 , primaryClass =

    Pith/arXiv arXiv

  72. [72]

    , keywords =

    A Family Tree of Optical Transients from Narrow-line Seyfert 1 Galaxies. , keywords =. doi:10.3847/1538-4357/ac110f , archivePrefix =. 2010.08554 , primaryClass =

    Pith/arXiv arXiv 2010

  73. [73]

    , keywords =

    The FIRST Survey: Faint Images of the Radio Sky at Twenty Centimeters. , keywords =. doi:10.1086/176166 , adsurl =

    doi:10.1086/176166

  74. [74]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2022-03-05. TNS Classification Report , keywords =

    2022

  75. [75]

    , keywords =

    A radio-emitting outflow produced by the tidal disruption event AT2020vwl. , keywords =. doi:10.1093/mnras/stad1258 , archivePrefix =. 2304.12661 , primaryClass =

    Pith/arXiv arXiv

  76. [76]

    , keywords =

    Tidal Disruption Event Demographics with the Zwicky Transient Facility: Volumetric Rates, Luminosity Function, and Implications for the Local Black Hole Mass Function. , keywords =. doi:10.3847/2041-8213/acf216 , archivePrefix =. 2303.06523 , primaryClass =

    Pith/arXiv arXiv 2041

  77. [77]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2021-05-19. TNS Classification Report , keywords =

    2021

  78. [78]

    , keywords =

    The Final Season Reimagined: 30 Tidal Disruption Events from the ZTF-I Survey. , keywords =. doi:10.3847/1538-4357/aca283 , archivePrefix =. 2203.01461 , primaryClass =

    Pith/arXiv arXiv

  79. [79]

    TNS Classification Report , keywords =

    ZTF Transient Classification Report for 2022-04-09. TNS Classification Report , keywords =

    2022

  80. [80]

    TNS Discovery Report , keywords =

    ZTF Transient Discovery Report for 2022-10-04. TNS Discovery Report , keywords =

    2022

Showing first 80 references.