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

2025-02-25

AI Summary

Astronomical transient classification faces two core challenges: poor cross-domain generalization (e.g., simulation-to-real, ZTF-to-LSST) and high annotation costs. This paper presents the first systematic validation of transfer learning for hierarchical astronomical time-series classification, proposing a lightweight deep neural network–based transfer framework. It aligns temporal features to accommodate heterogeneous sampling rates and noise characteristics across domains and introduces an efficient fine-tuning strategy. Experiments demonstrate that, on real ZTF data, the method reduces annotation requirements by 75% without sacrificing classification accuracy; on LSST simulation tasks, it achieves 95% of baseline accuracy using only 30% of the training data. The approach significantly shortens the deployment timeline for new survey classification models—particularly critical during early LSST operations—and establishes a transferable, low-annotation paradigm for large-scale astronomical time-series analysis.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

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 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 75% while maintaining equivalent performance to models trained from scratch. Similarly, when adapting ZTF models for LSST simulations, transfer learning achieves 95% 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.

Problem

• Transfer learning for astronomical transient classification
• Overcoming simulation-to-real data transfer challenges
• Adapting models from ZTF to LSST surveys

Innovation

• Transfer Learning application
• Reduced training data need
• Enhanced model adaptability