MiraBest

MiraBest数据集是由曼彻斯特大学物理与天文学院的Fiona A. M. Porter和Anna M. M. Scaife创建的，包含1256个经过形态分类的射电星系数据，用于机器学习应用。该数据集从NVSS和FIRST调查中筛选出0.03 < A? < 0.1的射电噪AGN，并由Miraghaei和Best（2017）手动标记，根据Fanaroff-Riley形态分类。MiraBest旨在支持深度学习库的使用，并解决天文学中不同机器学习算法性能评估的标准化数据集需求。数据集的应用领域包括射电星系的形态分类和机器学习模型的训练与评估。

The MiraBest dataset was created by Fiona A. M. Porter and Anna M. M. Scaife from the School of Physics and Astronomy, University of Manchester. It contains 1,256 morphologically classified radio galaxy data for machine learning applications. This dataset selects radio-loud AGN with 0.03 < A? < 0.1 from the NVSS and FIRST surveys, and was manually annotated by Miraghaei and Best (2017) based on the Fanaroff-Riley morphological classification. The MiraBest dataset aims to support the utilization of deep learning libraries and address the demand for standardized datasets for performance evaluation of various machine learning algorithms in astronomy. Its application fields include morphological classification of radio galaxies as well as the training and evaluation of machine learning models.