Explaining deep learning of galaxy morphology with saliency mapping

Abstract

We successfully demonstrate the use of explainable artificial intelligence (XAI) techniques on astronomical datasets in the context of measuring galactic bar lengths. The method consists of training convolutional neural networks on human classified data from Galaxy Zoo in order to predict general galaxy morphologies, and then using SmoothGrad (a saliency mapping technique) to extract the bar for measurement by a bespoke algorithm. We contrast this to another method of using a convolutional neural network to directly predict galaxy bar lengths. These methods achieved correlation coefficients of 0.76 and 0.59, and root mean squared errors of 1.69 and 2.10 respective to human measurements. We conclude that XAI methods outperform conventional deep learning in this case, which could be reasonably explained by the larger datasets available when training the models. We suggest that our XAI method can be used to extract other galactic features (such as the bulge-to-disk ratio) without needing to collect new datasets or train new models. We also suggest that these techniques can be used to refine deep learning models as well as identify and eliminate bias within training datasets.

arXiv comment11 pages, 5 figures, 2 tables, 2 appendices. Accepted for publication by Monthly Notices of the Royal Astronomical Society

References in corpus (26)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
• Striving for Simplicity: The All Convolutional Net
• Secular Evolution and the Formation of Pseudobulges in Disk Galaxies
• The Sloan Digital Sky Survey Photometric Camera
• Galaxy Zoo : Morphologies derived from visual inspection of galaxies from the Sloan Digital Sky Survey
• Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• SmoothGrad: removing noise by adding noise
• Galaxy Zoo 2: detailed morphological classifications for 304,122 galaxies from the Sloan Digital Sky Survey
• Improving galaxy morphologies for SDSS with Deep Learning
• Xception: Deep Learning with Depthwise Separable Convolutions
• Identity Mappings in Deep Residual Networks
• Galaxy Zoo DECaLS: Detailed Visual Morphology Measurements from Volunteers and Deep Learning for 314,000 Galaxies
• Galaxy Zoo: Reproducing Galaxy Morphologies Via Machine Learning
• Galaxy Zoo: Probabilistic Morphology through Bayesian CNNs and Active Learning
• Rings and spirals in barred galaxies. II. Ring and spiral morphology
• Neural computation as a tool for galaxy classification : methods and examples
• Galaxies, Human Eyes and Artificial Neural Networks
• Galaxy Zoo: Bar Lengths in Nearby Disk Galaxies
• Transfer learning for galaxy morphology from one survey to another
• Removing Astrophysics in 21 cm maps with Neural Networks
• Benchmarking and Scalability of Machine Learning Methods for Photometric Redshift Estimation
• Interpreting automatic AGN classifiers with saliency maps

Cited by in corpus (10)

• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• Galaxy Zoo: Kinematics of strongly and weakly barred galaxies
• Galaxy Morphology Classification using Neural Ordinary Differential Equations
• Unsupervised Galaxy Morphological Visual Representation with Deep Contrastive Learning
• Machine learning technique for morphological classification of galaxies from the SDSS. III. Image-based inference of detailed features
• Psi-GAN: A power-spectrum-informed generative adversarial network for the emulation of large-scale structure maps across cosmologies and redshifts
• Massive $\nu$s through the CNN lens: interpreting the field-level neutrino mass information in weak lensing
• Interpreting Cosmological Information from Neural Networks in the Hydrodynamic Universe
• Interpreting anomaly detection of SDSS spectra
• COSMOS2025: Machine Learning Classification of Early- and Late-type Galaxies at 0 < z < 3