An Early Classification Algorithm for Small Sample Transient Source Based on Machine Learning

Mengci Li; Chengzhi Liu; Chao Wu; Zhe Kang; Shiyu Deng; Zhenwei Li

doi:10.3788/AOS241045

Acta Optica Sinica, Volume. 44, Issue 24, 2430005(2024)

An Early Classification Algorithm for Small Sample Transient Source Based on Machine Learning

Mengci Li^1,2, Chengzhi Liu^1,2,3、*, Chao Wu^2,4、**, Zhe Kang^1,2, Shiyu Deng^1,4, and Zhenwei Li^1,2

¹Changchun Observatory, National Astronomical Observatories, Chinese Academy of Sciences, Changchun 130117, Jilin , China

²University of Chinese Academy of Sciences, Beijing 100049, China

³Key Laboratory of Space Object & Debris Observation, Purple Mountain Observatory, Chinese Academy of Sciences, Nanjing 210008, Jiangsu , China

⁴National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(9)

Fig. 1. Model structure of TXW algorithm

Download full size

Fig. 2. Example of transients. (a) TDE; (b) KN; (c) SNIa; (d) SLSN-I

Download full size

Fig. 3. Experimental process

Download full size

Fig. 4. Example of classification results. (a) Modified light curves of object; (b) probability of predicted object category at each time point

Download full size

Fig. 5. Confusion matrix. (a) Rapid algorithm at 2 d since trigger; (b) Rapid algorithm at 24 d since trigger; (c) TXW algorithm at 2 d since trigger; (d) TXW algorithm at 24 d since trigger

Download full size

Fig. 6. PR curves. (a) Rapid algorithm at 2 d since trigger; (b) Rapid algorithm at 24 d since trigger; (c) TXW algorithm at 2 d since trigger; (d) TXW algorithm at 24 d since trigger

Download full size

Fig. 7. ROC curves. (a) Rapid algorithm at 2 d since trigger; (b) Rapid algorithm at 24 d since trigger; (c) TXW algorithm at 2 d since trigger; (d) TXW algorithm at 24 d since trigger

Download full size

Fig. 8. AUC-time curves. (a) Rapid algorithm; (b) TXW algorithm

Download full size

Table 1. Comparison of classification accuracy of different algorithms
View table
Table 1. Comparison of classification accuracy of different algorithms
Method LSTM Transformer CNN RF Rapid TXW （no Weight） TXW
Accuracy /% 62.10 65.85 67.49 72.88 79.75 83.27 84.08

Tools

Get Citation

Copy Citation Text

Mengci Li, Chengzhi Liu, Chao Wu, Zhe Kang, Shiyu Deng, Zhenwei Li. An Early Classification Algorithm for Small Sample Transient Source Based on Machine Learning[J]. Acta Optica Sinica, 2024, 44(24): 2430005

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites