Oral Contribution (O4.2) Yanxia Zhang (National Astronomical Observatory,Chinese Academy of Sciences)
Theme: Data science challenges: tools from statistics to machine learning
Photometric Redshift Estimation of Quasars by Machine Learning
We summarize various techniques for photometric redshift estimation of quasars. Based on various datasets from different survey databases, we compare many machine learning algorithms on this issue. In reality, the more complex algorithms maybe not achieve better performance. The performance of an algorithm is influenced by different factors, such as feature selection, sample size and so on. In addition we also put forward a new strategy to improve the accuracy of photometric redshift estimation of quasars. The result shows that this strategy is applicable and efficient for our case. In future, we will apply it on the new survey databases, such as LSST.
