Kaydedildi:
Detaylı Bibliyografya
Yazar: Zhang, Jincheng
Materyal Türü: Recurso digital
Dil:
Baskı/Yayın Bilgisi: Zenodo 2025
Online Erişim:https://doi.org/10.5281/zenodo.17845170
Etiketler: Etiketle
Etiket eklenmemiş, İlk siz ekleyin!
İçindekiler:
  • <p><span>With the advent of the big data era in modern astronomy, astronomical observation projects have generated massive amounts of multimodal and multi-band data. Traditional optimization methods often struggle to achieve efficient, globally optimal solutions in high-dimensional, sparse, and nonlinear data analysis problems. Therefore, designing a heuristic optimization algorithm that incorporates the characteristics of astronomical data science is particularly important. This paper proposes an Astro-Informed Multi-scale Optimization (AIMO) algorithm, which innovatively introduces three mechanisms: astronomical data feature-driven optimization, multi-scale gravity-photometric fusion, and sparse event adaptive perturbation. The algorithm uses pure mathematical formulas to characterize the update process of the search particle in multi-band space, balancing global exploration and local exploitation while enhancing its responsiveness to sparse events. This paper details the algorithm structure, core mathematical formulas, and mechanism design, providing a theoretical foundation for the optimization and analysis of astronomical big data.</span></p>