miniJPAS与J-NEP联合数据集
收藏资源简介:
本研究构建的miniJPAS与J-NEP联合数据集是由J-PAS巡天路径探测器采集的多波段天文观测数据集,覆盖北黄极与AEGIS天区共计1.29平方度天区。数据集包含14,594个经过严格交叉认证的天体源,涵盖60个光学波段的光度测量数据与形态学参数,数据来源整合了SDSS、Gaia、DESI等六大权威巡天项目的光谱与形态分类标签。通过自动化质量控制流程与多源数据融合技术,构建过程确保了样本的纯净度与空间覆盖完整性。该数据集主要应用于星系-恒星分类的机器学习模型训练,旨在解决传统形态学分类在暗弱天体识别中的局限性,为宇宙学大尺度结构研究提供可靠的数据支撑。
The joint miniJPAS and J-NEP dataset constructed in this study is a multi-band astronomical observation dataset collected by the J-PAS survey pathfinder telescope, covering a total of 1.29 square degrees of sky across the North Ecliptic Pole and the AEGIS field. The dataset contains 14,594 rigorously cross-identified astronomical sources, including photometric measurement data across 60 optical bands and morphological parameters. It integrates spectral and morphological classification labels from six authoritative sky survey projects such as SDSS, Gaia, DESI, and others. Through automated quality control workflows and multi-source data fusion technologies, the dataset construction ensures the sample purity and spatial coverage completeness. This dataset is mainly used for training machine learning models for galaxy-star classification, aiming to address the limitations of traditional morphological classification in identifying faint celestial objects, and provide reliable data support for cosmological large-scale structure research.

- 1The miniJPAS and J-NEP surveys: Machine learning for star-galaxy separation巴西圣埃斯皮里图联邦大学 · 2025年



