[Data mining approach to cataclysmic variables candidates based on random forest algorithm].
[Data mining approach to cataclysmic variables candidates based on random forest algorithm].
复制标题
DOI：
--
复制
发表时间：
2012-02
期刊：
影响因子：
--
通讯作者：
Bin Jiang;A. Luo;Yongheng Zhao
中科院分区：
文献类型：
--
作者：
Bin Jiang;A. Luo;Yongheng Zhao
本文提出了一种自动有效的激变变量候选方法。选择识别的CV作为模板。采用随机森林算法，通过模板和随机选择光谱来构建模型。该模型描述了波长排序，并在此基础上构造了分类器。大多数非候选人都被排除在外。模板匹配策略用于识别最终候选人，并对这些候选人进行分析，以补充模板作为反馈。在实验中发现了16个新的CV候选者，表明我们的方法在LAMOST中寻找特殊天体是可行的。
An automatic and efficient method for cataclysmic variables candidates is presented in the present paper. The identified CVs were selected as templates. A model was constructed by random forest algorithm with templates and random selected spectra. Wavelength ranking was described by the model and the classifier was constructed afterwards. Most of the non-candidates were excluded by the method. Template matching strategy was used to identify the final candidates which were analyzed to complement the templates as feedback. 16 new CVs candidates were found in the experiment that shows that our approach to finding special celestial bodies can be feasible in LAMOST.