Adaptive stellar spectral subclass classification based on Bayesian SVMs
详细信息 查看全文
- 作者：Changde Dua ; b ; Ali Luo ; a ; lal@bao.ac.cn" class="auth_mail" title="E-mail the corresponding author ; lal@nao.cas.cn" class="auth_mail" title="E-mail the corresponding author ; Haifeng Yanga ; b ; c
- 关键词：Stellar spectral classification ; Bayesian support vector machines ; MCMC ; Data-driven
- 刊名：New Astronomy
- 出版年：2017
- 出版时间：February 2017
- 年：2017
- 卷：51
- 期：Complete
- 页码：51-58
- 全文大小：1441 K
文摘
- •
We applied the BSVM model to the stellar spectral subclass classification problems.
- •
BSVM model can infer the model parameters automatically, i.e., it is data-driven.
- •
We provided the optimal hyper-parameters α and β for BSVM model.
- •
We investigated different normalization methods and the best one has been suggested.
- •
BSVM model owns favorable adaptability and provides good prediction performance.