Assessing the Performance of a Machine Learning Algorithm in Identifying Bubbles in Dust Emission

Abstract

Stellar feedback created by radiation and winds from massive stars plays a significant role in both physical and chemical evolution of molecular clouds. This energy and momentum leaves an identifiable signature ("bubbles") that affect the dynamics and structure of the cloud. Most bubble searches are performed "by-eye", which are usually time-consuming, subjective and difficult to calibrate. Automatic classifications based on machine learning make it possible to perform systematic, quantifiable and repeatable searches for bubbles. We employ a previously developed machine learning algorithm, Brut, and quantitatively evaluate its performance in identifying bubbles using synthetic dust observations. We adopt magneto-hydrodynamics simulations, which model stellar winds launching within turbulent molecular clouds, as an input to generate synthetic images. We use a publicly available three-dimensional dust continuum Monte-Carlo radiative transfer code, hyperion, to generate synthetic images of bubbles in three Spitzer bands (4.5 um, 8 um and 24 um). We designate half of our synthetic bubbles as a training set, which we use to train Brut along with citizen-science data from the Milky Way Project. We then assess Brut's accuracy using the remaining synthetic observations. We find that after retraining Brut's performance increases significantly, and it is able to identify yellow bubbles, which are likely associated with B-type stars. Brut continues to perform well on previously identified high-score bubbles, and over 10% of the Milky Way Project bubbles are reclassified as high-confidence bubbles, which were previously marginal or ambiguous detections in the Milky Way Project data. We also investigate the size of the training set, dust model, evolution stage and background noise on bubble identification.

ApJ accepted

References in corpus (31)

• Interstellar Dust Grains
• The Spitzer c2d Legacy Results: Star Formation Rates and Efficiencies; Evolution and Lifetimes
• GLIMPSE: I. A SIRTF Legacy Project to Map the Inner Galaxy
• Hi-GAL: the Herschel infrared Galactic Plane Survey
• Rosseland and Planck Mean Opacities for Protoplanetary Discs
• Outflow Feedback Regulated Massive Star Formation in Parsec-Scale Cluster Forming Clumps
• Bolocam Survey for 1.1 mm Dust Continuum Emission in the c2d Legacy Clouds. I. Perseus
• On the Role of Massive Stars in the Support and Destruction of Giant Molecular Clouds
• Bolocam Survey for 1.1 mm Dust Continuum Emission in the c2d Legacy Clouds. II. Ophiuchus
• Protostellar Turbulence Driven by Collimated Outflows
• Before the first supernova: combined effects of HII regions and winds on molecular clouds
• The Milky Way Project First Data Release: A Bubblier Galactic Disk
• HYPERION: An open-source parallelized three-dimensional dust continuum radiative transfer code
• The GALFA-HI Survey: Data Release 1
• Feedback from Winds and Supernovae in Massive Stellar Clusters. I: Hydrodynamics
• Photoionising feedback in star cluster formation
• The COMPLETE Survey of Outflows in Perseus
• The Spitzer c2d Survey of Large, Nearby, Interstellar Clouds: VI. Perseus Observed with MIPS
• A Detailed Study of Feedback from a Massive Star
• A Bubbling Nearby Molecular Cloud: COMPLETE Shells in Perseus
• Massive stars in massive clusters IV: Disruption of clouds by momentum-driven winds
• Disruption of Molecular Clouds by Expansion of Dusty H II Regions
• The effect of stellar winds on the formation of a protocluster
• The Milky Way Project: Leveraging Citizen Science and Machine Learning to Detect Interstellar Bubbles
• Outflows and Bubbles in Taurus: Star-formation Feedback Sufficient to Maintain Turbulence
• Impact of Winds from Intermediate-Mass Stars on Molecular Cloud Structure and Turbulence
• Automated Classification of Periodic Variable Stars detected by the Wide-field Infrared Survey Explorer
• Classifying structures in the ISM with Support Vector Machines: the G16.05-0.57 supernova remnant
• The Milky Way Project: What are Yellowballs?
• Insights from Synthetic Star-forming Regions: I. Reliable Mock Observations from SPH Simulations
• Observations of 6 - 200 μm emission of the Ophiuchus cloud LDN 1688

Cited by in corpus (10)

• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• Zooming in on Individual Star Formation: Low- and High-mass Stars
• The Milky Way Project Second Data Release: Bubbles and Bow Shocks
• Synthetic observations of star formation and the interstellar medium
• CASI: A Convolutional Neural Network Approach for Shell Identification
• Application of Convolutional Neural Networks to Identify Stellar Feedback Bubbles in CO Emission
• Application of Convolutional Neural Networks to Identify Protostellar Outflows in CO Emission
• Infrared bubble recognition in the Milky Way and beyond using deep learning
• Turbulent Action at a Distance due to Stellar Feedback in Magnetized Clouds
• Investigations of MWISP Bubbles: Identification and Analysis of Enclosed Molecular Bubbles by Weight Fields