X-ray Binary ML: 3 Machine Learning Methods for Classifying the Compact Objects in X-ray Binaries

What are X-ray Binaries? X-ray binaries (XRBs) consist of a compact object accreting matter from a main sequence or supergiant companion star, orbiting the common center of mass.

Code Author

Zoe de Beurs: @zdebeurs

We customized code from the LIBSVM library (Chang & Lin 2011), the class package (Venables & Ripley 2002), and the Kernlab library (Karatzoglou et al. 2004).

Background

This directory contains Machine Learning Algorithms (Bayesian Gaussian Process, K-Nearest Neighbors, Support Vector Machines) for classifying the compact objects in X-ray Binaries. For more details, see our live-broadcasted talk at the SAO Astronomy Summer Intern Symposium 2019.

Citation

If you find this code useful, please cite our poster or our paper:

de Beurs, Z. L., Islam, N., Gopalan, G., & Vrtilek, S.D. 2020, American Astronomical Society Meeting Abstracts, 170.12

or

de Beurs, Z. L., Islam, N., Gopalan, G., & Vrtilek, S.D. (2020). A Comparative Study of Machine Learning Methods for X-ray Binary Classification. Submitted to ApJ.

Figures from the paper in 3D Animations

To examine the 3D Color-Color-Intensity Diagrams, go to Figures

Walkthrough

Setup

First, ensure that you have installed required packages in R:

• ggplot2 (instructions)
• e1071 (instructions)
• doFuture (instructions)

---

< UNDER CONSTRUCTION >

** NEW CONTENT WILL BE ADDED SOON **