线性逼近类通用频率原则获理论证明

1. 13:00
   arXiv 机器学习

   线性逼近类的通用频率原则

   有限线性逼近类在梯度下降训练下的通用频率原则获理论证明：在合适假设下，高频处的学习器振幅相对于目标振幅在整个训练过程中均趋于零。研究进一步给出了低频优先的充分条件，指出在特定比较条件下高频分量通常比低频需要更长的学习时间。基于勒让德基与三角基的数值实验验证了该理论结论。

Current heat 9·Comparable peak 10（Nov 2 14:00）·Comparable change, last 24 hours–

Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.