arXiv subjects

Ik Siong Heng

Publications and source records attributed to Ik Siong Heng.

At least 19 recordsLinked to original sources

arXiv2026

Probabilistic kilonova prediction from gravitational wave inferred binary neutron star parameters

Kilonovae provide a key electromagnetic window into binary neutron star mergers, revealing the properties of the merging system while probing r-process nucleosynthesis of the Universe. However, only a few kilonova candidates have been detected to date, with AT2017gfo being the only one associated with a gravitational wave event, GW170817. In this work, we present \textsc{Genova}, a probabilistic framework for predicting kilonova spectra and light curves directly from gravitational wave posterior samples of binary neutron star mergers. This method uses a conditional normalising flow to learn the distribution of rest-frame spectra conditioned on the source-frame component masses, tidal deformabilities, viewing angle and time since merger. Other kilonova model parameters, such as ejecta opacities, are marginalised over during training, so that their effects are propagated into the predicted spectra as predictive uncertainty. In the self-consistency test, the flow model reproduces the median light curves with residuals typically below $\sim 0.1$ mag, and the ratio of the predicted central 68\% interval widths remains predominantly between $0.8$ and $1.4$ over $\sim 0.4$--$8.0$ days. Comparisons with a physically distinct kilonova model show that the probabilistic prediction can remain informative beyond the model used for training. We apply \textsc{Genova} to GW170817/AT2017gfo using multi-band observations, including newly re-reduced $Y$-, $J$-, $K_s$-band photometry from the Visible and Infrared Survey Telescope for Astronomy, which we present in this work. The resulting predictive intervals broadly encompass the observations while capturing both gravitational wave posterior uncertainty and the variation induced by marginalised kilonova model parameters.

astro-ph.HE↗

arXiv2025

Leveraging cross-detector parameter consistency measures to enhance sensitivities of gravitational wave searches

All-sky searches for generic short-duration astrophysical GW transients are often challenging because of noise transients. Developing novel signal-noise discriminators is crucial for GW transient searches with LIGO Scientific, Virgo, and KAGRA (LVK) detectors. In this work, we adapt a recently developed Jensen Shannon divergence (JSD)-based measure, which assesses the cross-detector parameter consistency to distinguish between weakly modeled or unmodelled astrophysical GW signals and loud noise triggers. We first extend a 2-detector JSD-based measure, developed in an earlier work, to a 3-detector network. We leverage this to modify the test statistic of the existing Coherent Waveburst (cWB)-Gaussian Mixture Modelling (GMM) algorithm for short-duration transients towards improving the search sensitivity to ad-hoc waveforms like Sine-Gaussians, Gaussian Pulses, and White Noise Bursts. We find that with the new method, which we term cWB-GMM-JSD, the sensitivity to the ad-hoc waveforms, given by $h_{\mathrm{rss50}}$, improves by $\sim 10-20 \%$ at an IFAR of 10 years for the 2-detector network consisting of LHO and LLO detectors, and by $\sim 5-10 \%$ at the same IFAR for the 3-detector network consisting of LHO, LLO and Virgo detectors. Finally, we apply the modified statistic in the revised data analysis pipeline on the publicly available data from the third observing run (O3) of the LIGO and Virgo detectors. Although we do not find any new event in the O3 data, we see a notable rise in the statistical significance of most of the known GW events, which further testifies to the enhancement in sensitivities.

gr-qc↗

arXiv2025

Using gravitational wave dark sirens to choose between host galaxy weighting models

Binary black hole (BBH) mergers,, an important source of gravitational-waves(GWs), are assumed to be hosted in galaxies. The probability of a galaxy to host a BBH is related to its properties, for example stellar mass and star formation rate. These properties can be estimated from observables, such as the luminosity in certain observation bands. We refer to this description of host galaxy properties as host galaxy weighting models. However, the host galaxy weighting model for BBHs has yet to be accurately determined. Population synthesis has provided a variety of host galaxy weighting models. Here we investigate whether it is possible to distinguish different host galaxy weighting models using a data driven approach. We use the GW cosmology tool gwcosmo with a simulated IGO-Virgo-KAGRA (LVK) fifth observing run (O5)-like observing scenario. We also construct a mock spectroscopic galaxy catalogue from MICECATv2. Our analysis compares the Bayes factors among three simple luminosity weighting models, r-band, g-band, and uniform weighting. The Bayes factors among different host galaxy weighting models show a minor preference for the true model for a ~200-detection case, and a decisive preference for the true model over the uniform model for a ~1000-detection case, which is strongly driven by a small number of well-localised events.

astro-ph.CO↗

arXiv2025

Search for dynamical black hole captures with Gaussian mixture modelling

Gravitational waves (GWs) are expected to originate from black holes interacting dynamically in dense astrophysical environments. In such environments, given that the velocity and cross section between the interacting black holes is low, dynamical capture may occur. Such events merge on very short timescales with high eccentricities and are expected to be detectable in the LIGO-Virgo-KAGRA (LVK) sensitivity band. In this work, we present a dedicated search for dynamical black hole capture events in the third LVK observing run with the coherent WaveBurst (cWB) algorithm enhanced with Gaussian mixture modeling (GMM) post-production. With this we consider two applications of GMM: a weakly-modeled approach searching for generic short transients under minimal assumptions, and a population informed approach, in which the GMM model is provided information on the parameter space occupied by the capture population. Although our search does not find any new significant GW events, we find that an informed GMM approach brings significant sensitivity improvements, enabling the detection of dynamical capture events up to a distance of 1.9 Gpc for a 200 $M_{\odot}$ equal mass binary. We present updated upper limit estimates of the rate at 90\% confidence, the most stringent of which is 0.15 Gpc$^{-3}$yr$^{-1}$, a 34\% improvement with respect to previous observational estimates. Furthermore, while the weakly-modeled GMM approach is less sensitive to dynamical capture systems, we find that it is possible for these events to be detected up to a distance of 1 Gpc in the cWB-GMM all-sky short search under minimal assumptions. Finally, with the confident detection of GW190521, we estimate the rate of similar events to be 0.94 Gpc$^{-3}$yr$^{-1}$, assuming the event originated from a dynamical capture.

gr-qc↗

arXiv2025

Efficacy of Image Similarity as a Metric for Augmenting Small Dataset Retinal Image Segmentation

Synthetic images are an option for augmenting limited medical imaging datasets to improve the performance of various machine learning models. A common metric for evaluating synthetic image quality is the Fréchet Inception Distance (FID) which measures the similarity of two image datasets. In this study we evaluate the relationship between this metric and the improvement which synthetic images, generated by a Progressively Growing Generative Adversarial Network (PGGAN), grant when augmenting Diabetes-related Macular Edema (DME) intraretinal fluid segmentation performed by a U-Net model with limited amounts of training data. We find that the behaviour of augmenting with standard and synthetic images agrees with previously conducted experiments. Additionally, we show that dissimilar (high FID) datasets do not improve segmentation significantly. As FID between the training and augmenting datasets decreases, the augmentation datasets are shown to contribute to significant and robust improvements in image segmentation. Finally, we find that there is significant evidence to suggest that synthetic and standard augmentations follow separate log-normal trends between FID and improvements in model performance, with synthetic data proving more effective than standard augmentation techniques. Our findings show that more similar datasets (lower FID) will be more effective at improving U-Net performance, however, the results also suggest that this improvement may only occur when images are sufficiently dissimilar.

eess.IV↗

arXiv2025

Decoding Long-duration Gravitational Waves from Binary Neutron Stars with Machine Learning: Parameter Estimation and Equations of State

Gravitational waves (GWs) from binary neutron stars (BNSs) offer valuable understanding of the nature of compact objects and hadronic matter, and the science potential will be greatly enhanced by the third-generation (3G) GW detectors, which are expected to detect BNS signals with order-of-magnitude improvements in duration, detection rates, and signal strength. However, the resulting computational demands for analyzing such prolonged signals pose a critical challenge that existing Bayesian methods cannot feasibly address in the 3G era. To bridge this critical gap, we demonstrate a machine learning-based workflow capable of producing source parameter estimation and constraints on equations of state (EOSs) for hours-long BNS signals in seconds with minimal hardware costs. We employ efficient compression of the GW data and EOS using neural networks, based on which we build normalizing flows for inference that can deliver results in seconds. The optimized computational cost of BNS signal analysis with our framework shows that machine learning has the potential to be an indispensable tool for future catalog-level BNS analyses, paving the way for large-scale investigations of BNS-related physics across the 3G observational landscape.

gr-qc↗

arXiv2025

Can Transformers help us perform parameter estimation of overlapping signals in gravitational wave detectors?

Overlapping signals represent one of the major data analysis challenges in next-generation gravitational wave detectors. We leverage Transformers and Normalizing Flows, state-of-the-art machine learning algorithms, to address the parameter estimation of overlapping binary black hole mergers in the Einstein Telescope (ET). Our proposed model combines a Transformer-based "Knowledge Extractor Neural Network" (KENN) with a Normalizing Flow (HYPERION) to perform rapid and unbiased inference over multiple overlapping black hole binary events. The choice of architecture leverages the strength of Transformers in capturing complex and long-range temporal structures in the strain time series data, while Normalizing Flows provide a powerful framework to sample posterior distributions. We demonstrate the effectiveness and robustness of our model over simulated gravitational wave signals, showing that it maintains the same level of accuracy regardless of the correlation level in the data. Moreover our model provides estimates of chirp mass and coalescence times within <10-20% from the true simulated value. The results obtained are promising and show how this approach might represent a first step toward a deep-learning based inference pipeline for ET and other future gravitational wave detectors.

gr-qc↗

arXiv2025

Strategic White Paper on AI Infrastructure for Particle, Nuclear, and Astroparticle Physics: Insights from JENA and EuCAIF

Artificial intelligence (AI) is transforming scientific research, with deep learning methods playing a central role in data analysis, simulations, and signal detection across particle, nuclear, and astroparticle physics. Within the JENA communities-ECFA, NuPECC, and APPEC-and as part of the EuCAIF initiative, AI integration is advancing steadily. However, broader adoption remains constrained by challenges such as limited computational resources, a lack of expertise, and difficulties in transitioning from research and development (R&D) to production. This white paper provides a strategic roadmap, informed by a community survey, to address these barriers. It outlines critical infrastructure requirements, prioritizes training initiatives, and proposes funding strategies to scale AI capabilities across fundamental physics over the next five years.

astro-ph.IM↗

arXiv2024

Applications of machine learning in gravitational wave research with current interferometric detectors

This article provides an overview of the current state of machine learning in gravitational-wave research with interferometric detectors. Such applications are often still in their early days, but have reached sufficient popularity to warrant an assessment of their impact across various domains, including detector studies, noise and signal simulations, and the detection and interpretation of astrophysical signals. In detector studies, machine learning could be useful to optimize instruments like LIGO, Virgo, KAGRA, and future detectors. Algorithms could predict and help in mitigating environmental disturbances in real time, ensuring detectors operate at peak performance. Furthermore, machine-learning tools for characterizing and cleaning data after it is taken have already become crucial tools for achieving the best sensitivity of the LIGO--Virgo--KAGRA network. In data analysis, machine learning has already been applied as an alternative to traditional methods for signal detection, source localization, noise reduction, and parameter estimation. For some signal types, it can already yield improved efficiency and robustness, though in many other areas traditional methods remain dominant. As the field evolves, the role of machine learning in advancing gravitational-wave research is expected to become increasingly prominent. This report highlights recent advancements, challenges, and perspectives for the current detector generation, with a brief outlook to the next generation of gravitational-wave detectors.

gr-qc↗

arXiv2024

Enhancing search pipelines for short gravitational wave transients with Gaussian mixture modelling

We present an enhanced method for the application of Gaussian Mixture Modelling (GMM) to the coherent WaveBurst (cWB) algorithm in the search for short-duration gravitational wave (GW) transients. The supervised Machine Learning method of GMM allows for the multi-dimensional distributions of noise and signal to be modelled over a set of representative attributes, which aids in the classification of GW signals against noise transients (glitches) in the data. We demonstrate that updating the approach to model construction eliminates bias previously seen in the GMM analysis, increasing the robustness and sensitivity of the analysis over a wider range of burst source populations. The enhanced methodology is applied to the generic burst all-sky short search in the LIGO-Virgo full third observing run (O3), marking the first application of GMM to the 3 detector Livingston-Hanford-Virgo network. For both 2- and 3- detector networks, we observe comparable sensitivities to an array of generic signal morphologies, with significant sensitivity improvements to waveforms in the low Quality factor parameter space at false alarm rates of 1 per 100 years. This proves that GMM can effectively mitigate blip glitches, which are one of the most problematic sources of noise for un-modelled GW searches. The cWB-GMM search recovers similar numbers of compact binary coalescence (CBC) events as other cWB post-production methods, and concludes on no new gravitational wave detection after known CBC events are removed.

gr-qc↗

arXiv2024

Detection, sky localization and early warning for binary neutron star mergers by detectors located in China of different configurations in third generation detector network

This work shows the results of an evaluation of the impact that a detector located in China, with a noise budget comparable to that of a proposed high-frequency detector with a 20 km arm length, an Einstein Telescope (ET) or a Cosmic Explorer (CE), could have on the network of ET-CE in terms of detection rate, localization, and providing early warning alert for simulated binary neutron star (BNS)s. The results indicate that a three-detector network including a Chinese detector could identify at least 4.4% more BNS mergers than an ET-CE network alone. The localization uncertainty could be reduced by a factor of more than 5 on average compared to the ET-CE network. With a three-detector network involving a Chinese detector, up to 89% of BNS mergers could be located within 10 square degrees of the sky 10 minutes prior to the merger. The assessment suggests that the potential for early warning signals is highest when the Chinese detector is similar to ET, whereas the sources are detected with the highest signal-to-noise ratio and localized to the smallest regions when the detector is more akin to CE. Interestingly, the C20N network (comprising ET+CE+C20) can achieve comparable localization performance as the ET network while outperforming the ETCN network (featuring the ET+CE+ an ET-like detector in China) in terms of detection capabilities, especially at large distances, indicating that adding a 20 km kilohertz detector in China to ET-CE network would make significant contributions at least as adding an ET-like detector in China to multi-messenger astronomy for almost all BNS observations.

astro-ph.CO↗

arXiv2024

Searching for gravitational-wave bursts with space-borne detectors

The millihertz gravitational wave band is expected to be opened by space-borne detectors like TianQin. Various mechanisms can produce short outbursts of gravitational waves, whose actual waveform can be hard to model. In order to identify such gravitational wave bursts and not to misclassify them as noise transients, we proposed a proof-of-principle energy excess method, that utilized the signal-insensitive channel to veto noise transients. We perform a test on simulated data, and for bursts with a signal-to-noise ratio of 20, even with the contamination of noise transient, our methods can reach a detection efficiency of 97.4% under a false alarm rate of once per year. However, more frequent occurrences of noise transients would lower the detection efficiency.

gr-qc↗

arXiv2024

Rapid neutron star equation of state inference with Normalising Flows

The first direct detection of gravitational waves from binary neutron stars on the 17th of August, 2017, (GW170817) heralded the arrival of a new messenger for probing neutron star astrophysics and provided the first constraints on neutron star equation of state from gravitational wave observations. Significant computational effort was expended to obtain these first results and therefore, as observations of binary neutron star coalescence become more routine in the coming observing runs, there is a need to improve the analysis speed and flexibility. Here, we present a rapid approach for inferring the neutron star equation of state based on Normalising Flows. As a demonstration, using the same input data, our approach, ASTREOS, produces results consistent with those presented by the LIGO-Virgo collaboration but requires < 1 sec to generate neutron star equation of state confidence intervals. Furthermore, ASTREOS allows for non-parametric equation of state inference. This rapid analysis will not only facilitate neutron star equation of state studies but can potentially enhance future alerts for electromagnetic follow-up observations of binary neutron star mergers.

gr-qc↗

arXiv2023

Search for gravitational-wave bursts in LIGO data at the Schenberg antenna sensitivity range

The Brazilian Mario Schenberg gravitational-wave detector remained operational until 2016 when it was disassembled. To assess the feasibility of reassembling the antenna, its capability to detect GW within its designed sensitivity parameters needs to be evaluated. Although the antenna is currently disassembled, insights can be gleaned from the O3 data of the LIGO detectors, given the similarities between Schenberg's ultimate sensitivity and the interferometers' sensitivity in the [3150-3260] Hz band. The search focused on signals lasting from ms to seconds, with no assumptions about their morphology, polarization, and arrival sky direction. Data analysis was performed using the coherent WaveBurst pipeline in the frequency range between 512 Hz and 4096 Hz, specifically targeting signals with bandwidths overlapping the Schenberg frequency band. However, the O3 data did not yield statistically significant evidence of GW bursts. This null result allowed for the characterization of the search efficiency in identifying simulated signal morphologies and setting upper limits on the GW burst event rate as a function of strain amplitude. The current search is sensitive to sources emitting isotropically $5\times10^{-6} M_{\odot}c^2$ in GWs from a distance of 10 kiloparsecs with a 50\% detection efficiency at a false alarm rate of 1 per 100 years. Moreover, we revisited estimations of detecting f-modes of neutron stars excited by glitches, setting the upper limit of the f-mode energy for the population of Galactic pulsars to $\sim 8 \times 10^{-8} M_{\odot}c^2$ at 3205 Hz. Our simulations suggest f-modes are an unlikely source of gravitational waves for the aSchenberg. Nevertheless, its potential in probing other types of GW short transients, such from giant flares from magnetars, post-merger phase of binary NSs, or the inspiral of binaries of primordial BHs with sub-solar masses, remains promising.

gr-qc↗

arXiv2023

Science with the Einstein Telescope: a comparison of different designs

The Einstein Telescope (ET), the European project for a third-generation gravitational-wave detector, has a reference configuration based on a triangular shape consisting of three nested detectors with 10 km arms, where in each arm there is a `xylophone' configuration made of an interferometer tuned toward high frequencies, and an interferometer tuned toward low frequencies and working at cryogenic temperature. Here, we examine the scientific perspectives under possible variations of this reference design. We perform a detailed evaluation of the science case for a single triangular geometry observatory, and we compare it with the results obtained for a network of two L-shaped detectors (either parallel or misaligned) located in Europe, considering different choices of arm-length for both the triangle and the 2L geometries. We also study how the science output changes in the absence of the low-frequency instrument, both for the triangle and the 2L configurations. We examine a broad class of simple `metrics' that quantify the science output, related to compact binary coalescences, multi-messenger astronomy and stochastic backgrounds, and we then examine the impact of different detector designs on a more specific set of scientific objectives.

gr-qc↗

arXiv2023

Unpacking merger jets: a Bayesian analysis of GW170817, GW190425 and electromagnetic observations of short gamma-ray bursts

We present a novel fully Bayesian analysis to constrain short gamma-ray burst jet structures associated with cocoon, wide-angle and simple top-hat jet models, as well as the binary neutron star merger rate. These constraints are made given the distance and inclination information from GW170817, observed flux of GRB170817A, observed rate of short gamma-ray bursts detected by Swift, and the neutron star merger rate inferred from LIGO's first and second observing runs. A separate analysis is conducted where a fitted short gamma-ray burst luminosity function is included to provide further constraints. The jet structure models are further constrained using the observation of GW190425 and we find that the assumption that it produced a GRB170817-like short gamma-ray burst that went undetected due to the jet geometry is consistent with previous observations. We find and quantify evidence for low luminosity and wide-angled jet structuring in the short gamma-ray burst population, independently from afterglow observations, with log Bayes factors of $0.45{-}0.55$ for such models when compared to a classical top-hat jet. Slight evidence is found for a Gaussian jet structure model over all others when the fitted luminosity function is provided, producing log Bayes factors of $0.25{-}0.9\pm0.05$ when compared to the other models. However without considering GW190425 or the fitted luminosity function, the evidence favours a cocoon-like model with log Bayes factors of $0.14\pm0.05$ over the Gaussian jet structure. We provide new constraints to the binary neutron star merger rates of $1{-}1300$Gpc$^{-3}$yr$^{-1}$ or $2{-}680$Gpc$^{-3}$yr$^{-1}$ when a fitted luminosity function is assumed.

astro-ph.HE↗

arXiv2022

Observational limits on the rate of radiation-driven binary black hole capture events

Dense astrophysical environments like globular clusters and galactic nuclei can host hyperbolic encounters of black holes which can lead to gravitational-wave driven capture. There are several astrophysical models which predict a fraction of binary black hole mergers to come from these radiation-driven capture scenarios. In this paper we present the sensitivity of a search towards gravitational-wave driven capture events for O3, the third observing run of LIGO and Virgo. We use capture waveforms produced by numerical relativity simulations covering four different mass ratios and at least two different values of initial angular momentum per mass ratio. We employed the most generic search for short-duration transients in O3 to evaluate the search sensitivity in this parameter space for a wide range in total mass in terms of visible spacetime volume. From the visible spacetime volume we determine for the first time the merger rate upper limit of such systems. The most stringent estimate of rate upper limits at 90\% confidence is $0.2~\mathrm{Gpc}^{-3}\,\mathrm{yr}^{-1}$ for an equal mass $200~M_\odot$ binary. Furthermore, in recent studies the event GW190521 has been suggested to be a capture event. With this interpretation of GW190521, we find the merger rate of similar events to be $0.47~\mathrm{Gpc}^{-3}\,\mathrm{yr}^{-1}$.

gr-qc↗

arXiv2022

Mimicking Mergers: Mistaking Black Hole Captures as Mergers

As the number of gravitational wave observations has increased in recent years, the variety of sources has broadened. Here we investigate whether it is possible for the current generation of detectors to distinguish between very short-lived gravitational wave signals from mergers between high-mass black holes, and the signal produced by a close encounter between two black holes which results in gravitational capture, and ultimately a merger. We compare the posterior probability distributions produced by analysing simulated signals from both types of progenitor events, both under ideal and realistic scenarios. We show that while, under ideal conditions it is possible to distinguish both progenitors, under more realistic conditions they are indistinguishable. This has important implications for the interpretation of such short signals, and we therefore advocate that these signals be the focus of additional investigation even when satisfactory results have been achieved from standard analyses.

gr-qc↗