A crucial yet challenging task in galaxy evolution studies is the identification of distant merging galaxies, a task that suffers from a variety of issues ranging from telescope sensitivities and limitations to the inherently chaotic morphologies of young galaxies. In this paper, we use random forests and convolutional neural networks to identify high-redshift JWST Cosmic Evolution Early Release Science Survey (CEERS) galaxy mergers. We train these algorithms on simulated 3 < z < 5 CEERS galaxies created from the IllustrisTNG subhalo morphologies and the Santa Cruz SAM light cone. We apply our models to observed CEERS galaxies at 3 < z < 5. We find that our models correctly classify ∼60%–70% of simulated merging and nonmerging galaxies; better performance on the merger class comes at the expense of misclassifying more nonmergers. We could achieve more accurate classifications, as well as test for a dependency on physical parameters such as gas fraction, mass ratio, and relative orbits, by curating larger training sets. When applied to real CEERS galaxies using visual classifications as ground truth, the random forests correctly classified 40%–60% of mergers and nonmergers at 3 < z < 4 but tended to classify most objects as nonmergers at 4 < z < 5 (misclassifying ∼70% of visually classified mergers). On the other hand, the CNNs tended to classify most objects as mergers across all redshifts (misclassifying 80%–90% of visually classified nonmergers). We investigate what features the models find most useful, as well as the characteristics of false positives and false negatives, and also calculate merger rates derived from the identifications made by the models.

中文翻译：

---

CEERS 关键文件。IX. 使用随机森林和卷积神经网络识别 CEERS NIRCam 图像中的星系合并


在星系演化研究中，一项关键但具有挑战性的任务是识别遥远的合并星系，这项任务受到各种问题的影响，从望远镜的敏感性和局限性到年轻星系固有的混乱形态。在本文中，我们使用随机森林和卷积神经网络来识别高红移 JWST 宇宙演化早期发布科学巡天 （CEERS） 星系合并。我们在模拟的 3 < z < 5 CEERS 星系上训练这些算法，这些星系由 IllustrisTNG 亚晕形态和圣克鲁斯 SAM 光锥创建。我们将我们的模型应用于观测到的 3 < z < 5 的 CEERS 星系。我们发现我们的模型正确分类了大约 60%-70% 的模拟合并和非合并星系;合并类的更好表现是以错误分类更多的非合并为代价的。我们可以通过策划更大的训练集来实现更准确的分类，并测试对物理参数（如气体分数、质量比和相对轨道）的依赖性。当使用视觉分类作为基本事实应用于真实的 CEERS 星系时，随机森林在 3 < z < 4 处正确地将 40%-60% 的合并和非合并分类，但在 4 < z < 5 处倾向于将大多数天体归类为非合并（错误分类了 ∼70% 的视觉分类合并）。另一方面，CNN 倾向于将大多数对象归类为跨所有红移的合并（错误分类了 80%-90% 的视觉分类非合并）。我们调查了模型认为最有用的特征，以及假阳性和假阴性的特征，还计算了从模型所做的鉴定得出的合并率。