StellarF: A Lora-Adapter Integrated Large Model Framework for Stellar Flare Forecasting with Historical & Statistical Data

2025-07-15

AI Summary

Stellar flare prediction suffers from sparse event samples and the absence of dedicated large models. This paper introduces the first parameter-efficient large model framework specifically designed for this task, integrating LoRA and Adapter techniques with two novel modules: a historical flare records module and a statistical features module, enabling multi-scale temporal pattern recognition. Evaluated on Kepler and TESS light-curve data using a newly constructed dataset, our method significantly outperforms existing approaches (F1-score improvement of 12.3%), demonstrating strong effectiveness, generalizability, and cross-disciplinary applicability. Key contributions include: (1) the first lightweight large-model paradigm tailored to stellar flare prediction; (2) a dual-path modeling mechanism that synergistically combines historical-event-driven learning with statistical prior knowledge; and (3) a scalable architecture supporting low-resource astronomical time-series forecasting.

Technology Category

Machine Learning: Large Multimodal Models (LMMs)Natural Language Processing: (Large) Language ModelsComputer Vision: Large Vision Models

Application Category

Web Mining and Content Analysis: Large pretrained models with web dataSearch and Retrieval-Augmented AI: Large language models for searchGraph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Stellar flare forecasting, a critical research frontier in astronomy, offers profound insights into stellar activity. However, the field is constrained by both the sparsity of recorded flare events and the absence of domain-specific large-scale predictive models. To address these challenges, this study introduces StellarF (Stellar Flare Forecasting), a novel large model that leverages Low-Rank (LoRA) and Adapter techniques to parameter-efficient learning for stellar flare forecasting. At its core, StellarF integrates an flare statistical information module with a historical flare record module, enabling multi-scale pattern recognition from observational data. Extensive experiments on our self-constructed datasets (derived from Kepler and TESS light curves) demonstrate that StellarF achieves state-of-the-art performance compared to existing methods. The proposed prediction paradigm establishes a novel methodological framework for advancing astrophysical research and cross-disciplinary applications.

Problem

• Forecasting stellar flares with sparse historical data
• Lack of domain-specific large-scale predictive models
• Multi-scale pattern recognition from observational data

Innovation

• Uses LoRA and Adapter for efficient learning
• Integrates statistical and historical flare data
• Achieves state-of-the-art performance on Kepler/TESS data