Speaker
Description
With fast growing data collected by the Fermi Large Area Telescope as a big data problem, manual classification has become an impossible task for astronomers. In this paper, we propose a novel framework using machine learning techniques with automatic feature selection algorithms for gamma-ray object classification. We automate parameter tuning rather than manual tuning used in some previous work. We found that using the Random Forest (RF) algorithm for feature selection can result in a better classification performance. Optimal results can be obtained for classifying AGNs/pulsars (accuracy > 98%) and young pulsars/millisecond pulsars (accuracy > 95%) by using the boosted logistic regression algorithm and RF as the classifier respectively. In comparison with the previous work by Saz Parkinson et al. (2016), our scheme leads to an improved performance.