Comparison of three Statistical Classification Techniques for Maser Identification
arXiv:1603.06395 · doi:10.1017/pasa.2016.13
Abstract
We applied three statistical classification techniques - linear discriminant analysis (LDA), logistic regression and random forests - to three astronomical datasets associated with searches for interstellar masers. We compared the performance of these methods in identifying whether specific mid-infrared or millimetre continuum sources are likely to have associated interstellar masers. We also discuss the ease, or otherwise, with which the results of each classification technique can be interpreted. Non-parametric methods have the potential to make accurate predictions when there are complex relationships between critical parameters. We found that for the small datasets the parametric methods logistic regression and LDA performed best, for the largest dataset the non-parametric method of random forests performed with comparable accuracy to parametric techniques, rather than any significant improvement. This suggests that at least for the specific examples investigated here accuracy of the predictions obtained is not being limited by the use of parametric models. We also found that for LDA, transformation of the data to match a normal distribution in the input parameters led to big improvements in accuracy. The different classification techniques had significant overlap in their predictions, further astronomical observations will enable the accuracy of these predictions to be tested.
17 pages (plus online only appendix), 10 figures, 9 tables Accepted for publication in Publications of the Astronomical Society of Australia on 11 March 2016 (ID: PAS-15112.R1)
References in corpus (23)
- GLIMPSE: I. A SIRTF Legacy Project to Map the Inner Galaxy
- Interpreting Spectral Energy Distributions from Young Stellar Objects. II. Fitting observed SEDs using a large grid of pre-computed models
- Interpreting Spectral Energy Distributions from Young Stellar Objects. I. A grid of 200,000 YSO model SEDs
- Spitzer Survey of the Large Magellanic Cloud, Surveying the Agents of a Galaxy's Evolution (SAGE) I: Overview and Initial Results
- The Bolocam Galactic Plane Survey: Survey Description and Data Reduction
- Science With The Australian Square Kilometre Array Pathfinder
- Methanol masers : Reliable tracers of the early stages of high-mass star formation
- 12.2-GHz methanol masers associated with 1.2-mm dust clumps: Quantifying high-mass star formation evolutionary schemes
- High and Intermediate-Mass Young Stellar Objects in the Large Magellanic Cloud
- A 95 GHz Class I Methanol Maser Survey Toward GLIMPSE Extended Green Objects (EGOs)
- Detection of new sources of methanol emission at 95 GHz with the Mopra telescope
- Infall and outflow within 400 AU from a high-mass protostar. 3-D velocity fields from methanol and water masers in AFLG 5142
- The Giant Molecular Cloud associated with RCW 106 -- A 1.2 mm continuum mapping study
- Fermi's Sibyl: Mining the gamma-ray sky for dark matter subhaloes
- How to Find More Supernovae with Less Work: Object Classification Techniques for Difference Imaging
- The Relationship between Class I and Class II Methanol Masers
- Molecular line mapping of the giant molecular cloud associated with RCW 106 - III. Multi-molecular line mapping
- Constraining the properties of 1.2-mm dust clumps that contain luminous water masers
- Masers associated with high-mass star formation regions in the Large Magellanic Cloud
- A search for 22-GHz water masers within the giant molecular cloud associated with RCW106
- A 95 GHz Class I Methanol Maser Survey Toward A Sample of GLIMPSE Point Sources Associated with BGPS Clumps
- Rapid, Machine-Learned Resource Allocation: Application to High-redshift GRB Follow-up
- G10.472+0.027: An Extreme water maser outflow associated with a Massive Protostellar Cluster