HAYATE：通过混合机器学习与模板拟合进行光度红移估计
HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting
AI总结：
本研究提出融合模板拟合与机器学习的HAYATE光度红移估计方法，通过人工红移生成训练数据，在宽红移范围实现了比EAZY更精准、更稳健且速度快约100倍的估计，还可通过迁移学习进一步提升性能。
摘要：
直接基于光谱红移训练的机器学习光度红移（photo-z）方法是传统模板拟合方法的可行替代方案，但在偏离训练集分布的新数据上泛化能力可能不佳。本研究提出一种用于宽范围光度红移估计的混合算法——结合人工神经网络与模板拟合的HAYATE（Hybrid Algorithm for WI(Y)de-range photo-z estimation with Artificial neural networks and TEmplate fitting），这是一种融合模板拟合与数据驱动方法的新型光度红移方法，其训练损失针对红移点估计和概率分布两方面进行优化。我们利用z<1.3的低红移星系光谱能量分布（SEDs）生成人工训练数据，将其人工红移至z=5。我们在ZFOURGE巡天数据上测试了该模型，证明在光谱完备性不足的宽红移范围之外，HAYATE可作为EAZY的可靠仿真器。该网络在初始低红移区域（z<1.3）实现了比EAZY更小的误差（σ_NMAD），获得了精确的光度红移估计；即使在高红移外推区域（1.3<z<5），其性能也与EAZY相当。同时，它提供了比EAZY更稳健的光度红移估计，异常值率更低（η_0.2≲1%），且运行速度比原始模板拟合方法快约100倍。我们还证明HAYATE提供了更可靠的红移概率分布函数（PDFs），其概率积分变换得分的分布比EAZY更平坦。通过使用光谱红移样本进行迁移学习，性能得到进一步提升。我们预计未来的大型巡天将受益于这种适用于宽红移范围观测的新型方法。
英文摘要：
Machine learning photo-z methods, trained directly on spectroscopic redshifts, provide a viable alternative to traditional template fitting methods but may not generalise well on new data that deviates from that in the training set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo-z estimation with Artificial neural networks and TEmplate fitting (HAYATE), a novel photo-z method that combines template fitting and data-driven approaches and whose training loss is optimised in terms of both redshift point estimates and probability distributions. We produce artificial training data from low-redshift galaxy SEDs at z<1.3, artificially redshifted up to z=5. We test the model on data from the ZFOURGE surveys, demonstrating that HAYATE can function as a reliable emulator of EAZY for the broad redshift range beyond the region of sufficient spectroscopic completeness. The network achieves precise photo-z estimations with smaller errors ($σ_{NMAD}$) than EAZY in the initial low-z region (z<1.3), while being comparable even in the high-z extrapolated regime (1.3<z<5). Meanwhile, it provides more robust photo-z estimations than EAZY with the lower outlier rate ($η_{0.2}\lesssim 1\%$) but runs $\sim100$ times faster than the original template fitting method. We also demonstrate HAYATE offers more reliable redshift PDFs, showing a flatter distribution of Probability Integral Transform scores than EAZY. The performance is further improved using transfer learning with spec-z samples. We expect that future large surveys will benefit from our novel methodology applicable to observations over a wide redshift range.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
