论文标题
Kilonovanet：具有条件变异自动编码器的Kilonova光谱的替代模型
KilonovaNet: Surrogate Models of Kilonova Spectra with Conditional Variational Autoencoders
论文作者
论文摘要
Kilonova光谱的详细辐射转移模拟在多通用天体物理学中起着至关重要的作用。在参数推理研究中，使用仿真结果需要从模拟输出中构建替代模型，以便在需要采样的算法中使用。在这项工作中，我们提出了Kilonovanet，这是用于建造Kilonova Spectra替代模型的条件变异自动编码器（CVAE）的实施。该方法可以直接在光谱上进行训练，从而消除了预处理光谱的开销时间，并大大加快了参数推理时间。我们构建了三个最先进的Kilonova模拟数据集的替代模型，并目前是深入的替代错误评估方法，通常可以将其应用于任何替代施工方法。通过从光谱替代物中创建合成光度观测，我们对观察到的GW170817的光曲线数据进行参数推断，并将结果与以前的分析进行比较。鉴于Kilonovanet在参数推断期间执行的速度，它将成为未来引力波观察跑步的有用工具，以快速分析潜在的Kilonova候选者
Detailed radiative transfer simulations of kilonova spectra play an essential role in multimessenger astrophysics. Using the simulation results in parameter inference studies requires building a surrogate model from the simulation outputs to use in algorithms requiring sampling. In this work, we present KilonovaNet, an implementation of conditional variational autoencoders (cVAEs) for the construction of surrogate models of kilonova spectra. This method can be trained on spectra directly, removing overhead time of pre-processing spectra, and greatly speeds up parameter inference time. We build surrogate models of three state-of-the-art kilonova simulation data sets and present in-depth surrogate error evaluation methods, which can in general be applied to any surrogate construction method. By creating synthetic photometric observations from the spectral surrogate, we perform parameter inference for the observed light curve data of GW170817 and compare the results with previous analyses. Given the speed with which KilonovaNet performs during parameter inference, it will serve as a useful tool in future gravitational wave observing runs to quickly analyze potential kilonova candidates
