Transfer Learning for Transient Classification: From Simulations to Real Data and ZTF to LSST

Abstract

Machine learning has become essential for automated classification of astronomical transients, but current approaches face significant limitations: classifiers trained on simulations struggle with real data, models developed for one survey cannot be easily applied to another, and new surveys require prohibitively large amounts of labelled training data. These challenges are particularly pressing as we approach the era of the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST), where existing classification models will need to be retrained using LSST observations. We demonstrate that transfer learning can overcome these challenges by repurposing existing models trained on either simulations or data from other surveys. Starting with a model trained on simulated Zwicky Transient Facility (ZTF) light curves, we show that transfer learning reduces the amount of labelled real ZTF transients needed by 95% while maintaining equivalent performance to models trained from scratch. Similarly, when adapting ZTF models for LSST simulations, transfer learning achieves 94% of the baseline performance while requiring only 30% of the training data. These findings have significant implications for the early operations of LSST, suggesting that reliable automated classification will be possible soon after the survey begins, rather than waiting months or years to accumulate sufficient training data.

7 pages, 5 figures, 1 table. Accepted for publication in MNRAS Letters. A version of this paper was also accepted to the Machine Learning for Astrophysics Workshop 2025 at ICML

References in corpus (23)

• Array Programming with NumPy
• The Astropy Project: Sustaining and Growing a Community-oriented Open-source Project and the Latest Major Release (v5.0) of the Core Package
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• The Zwicky Transient Facility: System Overview, Performance, and First Results
• The Zwicky Transient Facility: Data Processing, Products, and Archive
• SNANA: A Public Software Package for Supernova Analysis
• The Zwicky Transient Facility Bright Transient Survey. II. A Public Statistical Sample for Exploring Supernova Demographics
• SuperNNova: an open-source framework for Bayesian, Neural Network based supernova classification
• RAPID: Early Classification of Explosive Transients using Deep Learning
• Models and Simulations for the Photometric LSST Astronomical Time Series Classification Challenge (PLAsTiCC)
• Avocado: Photometric Classification of Astronomical Transients with Gaussian Process Augmentation
• Deep Recurrent Neural Networks for Supernovae Classification
• SuperRAENN: A Semi-supervised Supernova Photometric Classification Pipeline Trained on Pan-STARRS1 Medium Deep Survey Supernovae
• Supernova Photometric Classification Pipelines Trained on Spectroscopically Classified Supernovae from the Pan-STARRS1 Medium-Deep Survey
• GHOST: Using Only Host Galaxy Information to Accurately Associate and Distinguish Supernovae
• Classifying Supernovae Using Only Galaxy Data
• FLEET: A Redshift-Agnostic Machine Learning Pipeline to Rapidly Identify Hydrogen-Poor Superluminous Supernovae
• Deep Attention-Based Supernovae Classification of Multi-Band Light-Curves
• Real-Time Detection of Anomalies in Large-Scale Transient Surveys
• ATAT: Astronomical Transformer for time series And Tabular data
• The First Two Years of FLEET: an Active Search for Superluminous Supernovae
• A Classifier-Based Approach to Multi-Class Anomaly Detection for Astronomical Transients
• A Classifier-Based Approach to Multi-Class Anomaly Detection Applied to Astronomical Time-Series

Cited by in corpus (2)

• The ATLAS Virtual Research Assistant
• The automation of optical transient discovery and classification in Rubin-era time-domain astronomy