A Stellar Spectrum Classification Algorithm Based on CNN and LSTM Composite Deep Learning Model
LI Hao1, ZHAO Qing1*, CUI Chen-zhou2, FAN Dong-wei2, ZHANG Cheng-kui1, SHI Yan-cui1, WANG Yuan1
1. School of Artificial Intelligence, Tianjin University of Science and Technology, Tianjin 300457, China
2. National Astronomical Observatory，Chinese Academy of Sciences，Beijing 100012，China
Abstract：Stellar spectral classification is a significant research direction in astronomy. With the rapid development of technology, the stellar spectral data collected by large survey telescopes have reached terabytes or even petabytes, and the traditional classification methods can no longer meet the processing needs of such a vast amount of data. CNNs learn the local features of the data by convolution operations, remove redundant information, and compress the features by maximum pooling operations. However, since the fully-connected layer of the original CNN model lacks the function of long-range dependency mining, this problem can be solved by adding LSTM networks, which can extract important features and detect small differences in features through their unique three “gates” of long-range dependency mining capability. Therefore, this paper proposes a deep model based on the composite of CNN and LSTM for classifying stellar spectra in LAMOST DR8. This model can better learn the features of stellar spectra, which provides an important help for stellar evolution studies. To improve the convergence speed of the model, the common Z-Score normalization method is used to process the data. The model proposed in this paper achieved a classification accuracy of 94.56% in the F, G, and K classification experiments. Meanwhile, compared with the previously used RBM, PILDNN, PILDNN*, DBN, Inception v3, 1D-SSCNN, and LSTM methods, the results show that the method in this paper has a higher classification accuracy. In the ten-class experiments, the method in this paper achieves 97.35% accuracy. The results are better than the experimental results using only LSTM and 1D-SSCNN methods, and the training time is reduced by nearly ten times. The F1 score is used to illustrate the classification accuracy of each class of stellar spectra, and the F1 value of each type is above 0.9 in both the three-classification and ten-class experiments. Compared with the results of previous experiments in the literature, the results of this paper's model are better. With the confusion matrix results, it can be concluded that the model's accuracy in this paper is higher in the experiments with more spectral categories, and it can even reach 100%. In summary, the model based on the combination of CNN and LSTM proposed in this paper can effectively classify large-scale stellar spectral data and achieve excellent classification results.
Key words：LAMOST;Spectral classification of stars;CNN;LSTM;Z-Score standardization
李 浩，赵 青，崔辰州，樊东卫，张成奎，史艳翠，王 嫄. 基于CNN与LSTM复合深度模型的恒星光谱分类算法[J]. 光谱学与光谱分析, 2024, 44(06): 1668-1675.
LI Hao, ZHAO Qing, CUI Chen-zhou, FAN Dong-wei, ZHANG Cheng-kui, SHI Yan-cui, WANG Yuan. A Stellar Spectrum Classification Algorithm Based on CNN and LSTM Composite Deep Learning Model. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2024, 44(06): 1668-1675.
[1] Zhao G, Zhao Y H, Chu Y Q, et al. Research in Astronomy and Astrophysics, 2012, 12(7): 723.
[2] Cui X Q, Zhao Y H, Chu Y Q, et al. Research in Astronomy and Astrophysics, 2012, 12(9): 1197.
[3] Kheirdastan S, Bazarghan M. Astrophysics and Space Science, 2016, 361: 304.
[4] CAI Jiang-hui, YANG Hai-feng, ZHAO Xu-jun, et al(蔡江辉, 杨海峰, 赵旭俊, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2013, 33(1): 237.
[5] Wang K, Guo P, Luo A L. Monthly Notices of the Royal Astronomical Society, 2017, 465(4): 4311.
[6] XU Ting-ting, MA Chen-ye, ZHANG Jing-min, et al(许婷婷, 马晨晔, 张静敏, 等). Acta Astronomica Sinica(天文学报), 2019, 60(2): 15.
[7] Kuntzer T, Tewes M, Courbin F. Astronomy & Astrophysics, 2016, 591: A54.
[8] Dafonte C, Rodríguez A, Manteiga M, et al. Entropy, 2020, 22(5): 518.
[9] Li X R, Lin Y T, Qiu K B. Research in Astronomy and Astrophysics, 2019, 19(8): 111.
[10] Liu C, Cui W Y, Zhang B, et al. Research in Astronomy and Astrophysics, 2015, 15(8): 1137.
[11] Liu W, Zhu M, Dai C, et al. Monthly Notices of the Royal Astronomical Society, 2019, 483(4): 4774.
[12] HE Dong-yuan, LIU Wei, CAO Shuo, et al(何东远, 刘 伟, 曹 硕, 等). Journal of Beijing Normal University(Natural Science Edition)[北京师范大学学报(自然科学版)], 2020, 56(1): 37.
[13] Zhang J M, Ma C Y, Wang L, et al. Chinese Astronomy and Astrophysics, 2020, 44(3): 334.
[14] Lu Y, Qiu B, Xiang G, et al. Journal of Physics: Conference Series. IOP Publishing, 2020, 1626(1): 012016.
[15] Vilavicencio-Arcadia E, Navarro S G, Corral L J, et al. Mathematical Problems in Engineering, 2020, 2020: 1751932.
[16] Yoon K. Conference on Empirical Methods in Natural Language Processing , 2014.
[17] HE Li, YUAN Bin(何 丽, 袁 斌). Computer Technology and Development(计算机技术与发展), 2019, 29(11): 190.
