arXiv subjects

Siddharth Chaini

Publications and source records attributed to Siddharth Chaini.

9 recordsLinked to original sources

arXiv2026

NAPTIME: A Neural-Process Framework for Rubin Alert Classification

The Vera C. Rubin Observatory Legacy Survey of Space and Time will produce a high-volume stream of irregularly sampled multiband alerts for which spectroscopic confirmation will be available only for a small minority of sources. Tidal disruption events are rare phenomena that provide a direct probe of dormant massive black holes, but their light curves can be confused with nuclear variability and other transient subclasses. We present NAPTIME (Neural Astrophysical Photometric Transient Identification and Modeling Engine), a neural-process framework for photometric transient classification under sparse and partial observational context. NAPTIME models irregular multiband light curves directly, combining probabilistic light-curve reconstruction with classification and optional host-galaxy context, as well as photometric-redshift information. We evaluate on two simulated benchmarks: ELAsTiCC2, our primary Rubin-like broad-classification benchmark, and MALLORN, a photometry-only TDE-focused benchmark. On the 15-family ELAsTiCC2 task, the metadata-aware model reaches macro $\mathrm{F1} = 0.903$ and macro $\mathrm{AUROC} = 0.991$, while a matched photometry-only variant reaches 0.874 and 0.986. Viewed as a TDE-versus-rest ranking model, the classifier yields TDE average precision 0.985 with metadata and 0.979 without. Metadata is most valuable in the low-context regime. Using only the earliest 10\% of detected observations, macro F1 is $\sim$0.42 with metadata and $\sim$0.34 without it. On MALLORN, NAPTIME reaches macro $\mathrm{F1} = 0.693$ and macro $\mathrm{AUROC} = 0.958$. These results show that neural processes provide a practical probabilistic framework for Rubin-like transient classification and remain effective for TDE-focused candidate recovery.

astro-ph.IM↗

arXiv2026

Probabilistic Data-Driven Modelling of Astrophysical Transients: The Neural Process Family for Ultrafast and Class-Agnostic Light Curve Reconstruction with NightLANP

Astrophysical observations from Earth are subject to weather, environmental, and scientific constraints that lead to sparse, irregular light curves. On the eve of the Vera C. Rubin Observatory Legacy Survey of Space and Time, its dataset offers unprecedented opportunities for transient science. Yet a key challenge remains its cadence, sparse and irregular across six bands, limiting inference. Interpolation helps mitigate this, with Gaussian Processes the standard, but they struggle with cross-band correlations, require a priori kernel specification, and must be fit to each light curve individually, hence scaling poorly. Here, we introduce the neural process family for light curve reconstruction, combining the probabilistic framework of Gaussian Processes with the scalability of deep learning. By meta-learning on diverse simulated transients, Attentive Neural Processes shift the bulk of computation to training, enabling rapid, amortized inference with a class-agnostic model. Evaluated on realistic Rubin cadences across 15 transient classes, we show that even an unoptimized, out-of-the-box Attentive Neural Process consistently outperforms all benchmarks -- a suite of Gaussian Processes and neural networks -- on every tested metric, spanning regression quality, astrophysical feature recovery, and probabilistic calibration. Our model interpolates all bands simultaneously in microseconds, over four orders of magnitude faster than the next-best neural benchmark and five faster than Gaussian Processes, demonstrating the potential of neural processes for the nightly Rubin alert stream. Attentive Neural Processes avoid the overconfidence of standard neural networks and the underconfidence of Gaussian Processes, delivering sharp, well-calibrated uncertainties. This work establishes the neural process family as a scalable, probabilistic foundation for real-time transient science in the Rubin era.

astro-ph.IM↗

arXiv2026

Searching for Ultracool Dwarfs in Early LSST Data Products

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) promises to drastically accelerate the discovery of ultracool dwarfs (UCDs) over the course of its 10-year survey of the Southern Hemisphere. With the official start of LSST imminent, we showcase LSST's capabilities for discovering and characterizing UCDs using early commissioning data (Data Preview 1). The LSST photometric system at this stage remains poorly understood for faint UCDs. Thus, we begin by cross-matching Data Preview 1 against known UCD catalogs. We recover 1 known UCD from the Ultracool Sheet, 17 UCDs from the Dark Energy Survey, and 17 low mass stars from the Gaia Catalog of Nearby Stars. Using these known UCDs alongside recent spectroscopically-confirmed Euclid objects, we select 89 ultracool dwarf candidates in LSST fields, 17 of which are unique to this work. We present our candidates, a photometric temperature estimate, and discuss lessons learned from using early LSST data products. Finally, we turn to the future and predict potential UCD counts in upcoming LSST commissioning data (Data Preview 2), which is expected to be available to the Rubin community in 2026. Using synthetic populations of brown dwarfs, we forecast over 17,000 objects may be discovered and characterized in Data Preview 2. We predict that several hundred known objects and thousands of as-of-yet undiscovered UCDs may be detected in Data Preview 2 fields.

astro-ph.SR↗

arXiv2025

In Search of the Unknown Unknowns: A Multi-Metric Distance Ensemble for Out of Distribution Anomaly Detection in Astronomical Surveys

Distance-based methods involve the computation of distance values between features and are a well-established paradigm in machine learning. In anomaly detection, anomalies are identified by their large distance from normal data points. However, the performance of these methods often hinges on a single, user-selected distance metric (e.g., Euclidean), which may not be optimal for the complex, high-dimensional feature spaces common in astronomy. Here, we introduce a novel anomaly detection method, Distance Multi-Metric Anomaly Detection (DiMMAD), which uses an ensemble of distance metrics to find novelties. Using multiple distance metrics is effectively equivalent to using different geometries in the feature space. By using a robust ensemble of diverse distance metrics, we overcome the metric-selection problem, creating an anomaly score that is not reliant on any single definition of distance. We demonstrate this multi-metric approach as a tool for simple, interpretable scientific discovery on astronomical time series -- (1) with simulated data for the upcoming Vera C. Rubin Observatory Legacy Survey of Space and Time, and (2) real data from the Zwicky Transient Facility. We find that DiMMAD excels at out-of-distribution anomaly detection -- anomalies in the data that might be new classes -- and beats other state-of-the-art methods in the goal of maximizing the diversity of new classes discovered. For rare in-distribution anomaly detection, DiMMAD performs similarly to other methods, but may allow for improved interpretability. All our code is open source: DiMMAD is implemented within DistClassiPy: https://github.com/sidchaini/distclassipy/, while all code to reproduce the results of this paper is available here: https://github.com/sidchaini/dimmad/.

astro-ph.IM↗

arXiv2024

Validation of the Bond et al. (2010) SDSS-derived kinematic models for the Milky Way's disk and halo stars with Gaia Data Release 3 proper motion and radial velocity data

We validate the Bond et al. (2010) kinematic models for the Milky Way's disk and halo stars with Gaia Data Release 3 data. Bond et al. constructed models for stellar velocity distributions using stellar radial velocities measured by the Sloan Digital Sky Survey (SDSS) and stellar proper motions derived from SDSS and the Palomar Observatory Sky Survey astrometric measurements. These models describe velocity distributions as functions of position in the Galaxy, with separate models for disk and halo stars that were labeled using SDSS photometric and spectroscopic metallicity measurements. We find that the Bond et al. model predictions are in good agreement with recent measurements of stellar radial velocities and proper motions by the Gaia survey. In particular, the model accurately predicts the skewed non-Gaussian distribution of rotational velocity for disk stars and its vertical gradient, as well as the dispersions for all three velocity components. Additionally, the spatial invariance of velocity ellipsoid for halo stars when expressed in spherical coordinates is also confirmed by Gaia data at galacto-centric radial distances of up to 15 kpc.

astro-ph.GA↗

arXiv2024

Light Curve Classification with DistClassiPy: a new distance-based classifier

The rise of synoptic sky surveys has ushered in an era of big data in time-domain astronomy, making data science and machine learning essential tools for studying celestial objects. While tree-based models (e.g. Random Forests) and deep learning models dominate the field, we explore the use of different distance metrics to aid in the classification of astrophysical objects. We developed DistClassiPy, a new distance metric based classifier. The direct use of distance metrics is unexplored in time-domain astronomy, but distance-based methods can help make classification more interpretable and decrease computational costs. In particular, we applied DistClassiPy to classify light curves of variable stars, comparing the distances between objects of different classes. Using 18 distance metrics on a catalog of 6,000 variable stars across 10 classes, we demonstrate classification and dimensionality reduction. Our classifier meets state-of-the-art performance but has lower computational requirements and improved interpretability. Additionally, DistClassiPy can be tailored to specific objects by identifying the most effective distance metric for that classification. To facilitate broader applications within and beyond astronomy, we have made DistClassiPy open-source and available at https://pypi.org/project/distclassipy/.

astro-ph.IM↗

arXiv2022

Photometric identification of compact galaxies, stars and quasars using multiple neural networks

We present MargNet, a deep learning-based classifier for identifying stars, quasars and compact galaxies using photometric parameters and images from the Sloan Digital Sky Survey (SDSS) Data Release 16 (DR16) catalogue. MargNet consists of a combination of Convolutional Neural Network (CNN) and Artificial Neural Network (ANN) architectures. Using a carefully curated dataset consisting of 240,000 compact objects and an additional 150,000 faint objects, the machine learns classification directly from the data, minimising the need for human intervention. MargNet is the first classifier focusing exclusively on compact galaxies and performs better than other methods to classify compact galaxies from stars and quasars, even at fainter magnitudes. This model and feature engineering in such deep learning architectures will provide greater success in identifying objects in the ongoing and upcoming surveys, such as Dark Energy Survey (DES) and images from the Vera C. Rubin Observatory.

astro-ph.GA↗

arXiv2020

Thermodynamic Properties of Ice: A Monte Carlo Study

In this text, we implement a monte carlo algorithm to study thermodynamic properties of ice. Our program, written in Python, is open-sourced and available at https://github.com/AKnightWing/ColdAsIce. We develop a novel scheme to compute the residual entropy of a two dimensional lattice model of ice at absolute zero. A model of energetic ice is also considered and its thermodynamic properties are studied with temperature. We report a remarkable result for the energetic ice model, the presence of a phase transition at non zero temperatures.

cond-mat.stat-mech↗

arXiv2020

Astronomical Classification of Light Curves with an Ensemble of Gated Recurrent Units

With an ever-increasing amount of astronomical data being collected, manual classification has become obsolete; and machine learning is the only way forward. Keeping this in mind, the Large Synoptic Survey Telescope (LSST) Team hosted the Photometric LSST Astronomical Time-Series Classification Challenge (PLAsTiCC) in 2018. The aim of this challenge was to develop models that accurately classify astronomical sources into different classes, scaling from a limited training set to a large test set. In this text, we report our results of experimenting with Bidirectional Gated Recurrent Unit (GRU) based deep learning models to deal with time series data of the PLAsTiCC dataset. We demonstrate that GRUs are indeed suitable to handle time series data. With minimum preprocessing and without augmentation, our stacked ensemble of GRU and Dense networks achieves an accuracy of 76.243%. Data from astronomical surveys such as LSST will help researchers answer questions pertaining to dark matter, dark energy and the origins of the universe; accurate classification of astronomical sources is the first step towards achieving this. Our code is open-source and has been made available on GitHub here: https://github.com/AKnightWing/Astronomical-Classification-PLASTICC

astro-ph.IM↗