Science is currently at an age where there is more data than we know how to deal with. Machine learning (ML) is an emerging tool that is useful in drawing valuable science out of incomprehensibly large datasets, identifying complex trends in data that are otherwise overlooked. Moreover, ML can potentially enhance the quality and quantity of scientific data as it is collected. This paper explores how a new ML method can improve the rate of classification of rare Ic-BL supernovae (SNe). New parameters called magnitude rates were introduced to train ML models to identify SNe Ic-BL in large datasets. The same methodology was applied to a population of SN Ia transients to see if the methodology could be reproducible with another SN class. Three magnitudes, three time differences, two magnitude rates and the second derivative of these rates were calculated using the first three available photometric data points in a single filter. Initial investigations show that the Random Forest algorithm provides a strong foundation for the early classifications SNe Ic-BL and SNe Ia. Testing this model again on an unseen dataset shows that the model can identify upward of 13% of the total true SN Ic-BL population, significantly improving upon current methods. By implementing a dedicated observation campaign using this model, the number of SN Ic-BL classified and the quality of early-time data collected each year will see considerable growth in the near future.
Solar active regions (ARs) do not appear randomly but cluster along longitudinally warped toroidal bands ('toroids') that encode information about magnetic structures in the tachocline, where global-scale organization likely originates. Global MagnetoHydroDynamic Shallow-Water Tachocline (MHD-SWT) models have shown potential to simulate such toroids, matching observations qualitatively. For week-scale early prediction of flare-producing AR emergence, forward-integration of these toroids is necessary. This requires model initialization with a dynamically self-consistent MHD state-vector that includes magnetic, flow fields, and shell-thickness variations. However, synoptic magnetograms provide only geometric shape of toroids, not the state-vector needed to initialize MHD-SWT models. To address this challenging task, we develop PINNBARDS, a novel Physics-Informed Neural Network (PINN)-Based AR Distribution Simulator, that uses observational toroids and MHD-SWT equations to derive initial state-vector. Using Feb-14-2024 SDO/HMI synoptic map, we show that PINN converges to physically consistent, predominantly antisymmetric toroids, matching observed ones. Although surface data provides north and south toroids' central latitudes, and their latitudinal widths, they cannot determine tachocline field strengths, connected to AR emergence. We explore here solutions across a broad parameter range, finding hydrodynamically-dominated structures for weak fields (~2 kG) and overly rigid behavior for strong fields (~100 kG). We obtain best agreement with observations for 20-30 kG toroidal fields, and ~10 degree bandwidth, consistent with low-order longitudinal mode excitation. To our knowledge, PINNBARDS serves as the first method for reconstructing state-vectors for hidden tachocline magnetic structures from surface patterns; potentially leading to weeks ahead prediction of flare-producing AR-emergence.
The Laser Interferometer Space Antenna (LISA) data stream will inevitably contain gaps due to maintenance and environmental disturbances, introducing nonstationarities and spectral leakage that compromise standard frequency-domain likelihood evaluations. We present a scalable Simulation-Based Inference (SBI) framework capable of robust parameter estimation directly from gapped time-series data. We employ Flow Matching Posterior Estimation (FMPE) conditioned on a learned summary of the data, optimized through an end-to-end training strategy. To address the computational challenges of long-duration signals, we propose a dual-pathway summarizer architecture: a 1D Convolutional Neural Network (CNN) operating on the time domain for high precision, and a novel wavelet-based 2D CNN utilizing asymmetric, dilated kernels to achieve scalability for datasets spanning months. We demonstrate the efficacy of this framework on simulated Galactic Binary-like signals, showing that our joint training approach yields tighter, unbiased posteriors compared to two-stage reconstruction pipelines. Furthermore, we provide the first systematic comparison showing that FMPE offers superior stability and coverage calibration over conventional Normalizing Flows in the presence of severe data artifacts.