Identifying complex sources in large astronomical data using a coarse-grained complexity measure

Abstract

The volume of data that will be produced by the next generation of astrophysical instruments represents a significant opportunity for making unplanned and unexpected discoveries. Conversely, finding unexpected objects or phenomena within such large volumes of data presents a challenge that may best be solved using computational and statistical approaches. We present the application of a coarse-grained complexity measure for identifying interesting observations in large astronomical data sets. This measure, which has been termed apparent complexity, has been shown to model human intuition and perceptions of complexity. Apparent complexity is computationally efficient to derive and can be used to segment and identify interesting observations in very large data sets based on their morphological complexity. We show, using data from the Australia Telescope Large Area Survey, that apparent complexity can be combined with clustering methods to provide an automated process for distinguishing between images of galaxies which have been classified as having simple and complex morphologies. The approach generalizes well when applied to new data after being calibrated on a smaller data set, where it performs better than tested classification methods using pixel data. This generalizability positions apparent complexity as a suitable machine-learning feature for identifying complex observations with unanticipated features.

arXiv comment15 pages, 8 figures. New version to match that appearing in the journal. Appearing in PASP Focus on Machine Intelligence in Astronomy and Astrophysics

References in corpus (13)

• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• A catalog of visual-like morphologies in the 5 CANDELS fields using deep-learning
• Deep ATLAS radio observations of the CDFS-SWIRE field
• Results from the Supernova Photometric Classification Challenge
• Classifying Radio Galaxies with Convolutional Neural Network
• The weirdest SDSS galaxies: results from an outlier detection algorithm
• Deep Recurrent Neural Networks for Supernovae Classification
• A Package for the Automated Classification of Periodic Variable Stars
• ATLAS - I. Third Release of 1.4 GHz Mosaics and Component Catalogues
• Extragalactic Radio Continuum Surveys and the Transformation of Radio Astronomy
• A simple and robust method for automated photometric classification of supernovae using neural networks
• Discovering the Unexpected in Astronomical Survey Data
• Radio Galaxy Zoo: Machine learning for radio source host galaxy cross-identification

Cited by in corpus (8)

• Unveiling the rarest morphologies of the LOFAR Two-metre Sky Survey radio source population with self-organised maps
• Anomaly detection in Hyper Suprime-Cam galaxy images with generative adversarial networks
• A unique, ring-like radio source with quadrilateral structure detected with machine learning
• Foreword to the Focus Issue on Machine Learning in Astronomy and Astrophysics
• Density Based Outlier Scoring on Kepler Data
• Identifying anomalous radio sources in the EMU Pilot Survey using a complexity-based approach
• Planetary Complexity Revealed by the Joint Differential Entropy of Eigencolours
• Quantifying Radio Source Morphology