LUMOS用稀疏学习自动选特征剪模型，让科研人员轻松搞科学机器学习。
LUMOS: Democratizing SciML Workflows with L0-Regularized Learning for Unified Feature and Parameter Adaptation
- 用L0正则化同时选关键特征和剪冗余参数
- 平均减少71.45%参数量，推理速度提升6.4倍
- 适合无经验者快速搭建高效科学机器学习模型
科学机器学习（SciML）快速发展，但构建有效模型仍需大量先验知识和人工调参，尤其在特征选择与模型规模确定方面。本文提出LUMOS，一种基于L0正则化学习的端到端框架，统一实现特征选择与模型剪枝，降低对人工干预的依赖。通过半随机门控与重参数化技术，LUMOS在训练中动态筛选重要特征并剪除冗余参数，保持预测精度。我们在13个不同领域的SciML任务上评估，涵盖宇宙学与分子科学。结果表明，LUMOS平均实现71.45%的参数压缩和6.4倍的推理加速。分布式数据并行（DDP）训练最多支持8张GPU，验证了其可扩展性。
原文摘要 · Abstract (English)
The rapid growth of scientific machine learning (SciML) has accelerated discovery across diverse domains, yet designing effective SciML models remains a challenging task. In practice, building such models often requires substantial prior knowledge and manual expertise, particularly in determining which input features to use and how large the model should be. We introduce LUMOS, an end-to-end framework based on L0-regularized learning that unifies feature selection and model pruning to democratize SciML model design. By employing semi-stochastic gating and reparameterization techniques, LUMOS dynamically selects informative features and prunes redundant parameters during training, reducing the reliance on manual tuning while maintaining predictive accuracy. We evaluate LUMOS across 13 diverse SciML workloads, including cosmology and molecular sciences, and demonstrate its effectiveness and generalizability. Experiments on 13 SciML models show that LUMOS achieves 71.45% parameter reduction and a 6.4x inference speedup on average. Furthermore, Distributed Data Parallel (DDP) training on up to eight GPUs confirms the scalability of
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。