2025

苏剑林-梯度视角下的LoRA：简介、分析、猜测及推广

https://spaces.ac.cn/archives/9590spaces.ac.cn

NFL 理论 - 对于任何学习算法存在失效的分布 -> 在一定的先验知识的划定范围内训练模型

No free lunch theoremWikipedia

Machine Learning for Stock SelectionTaylor & Francis

How Good is Your Algorithm at Predicting Stock Prices Really?Medium