Why can a decaying learning rate slow adaptation to a genuinely new pattern late in a stream?
Classic MLLevel: AdvancedDecaying steps and drift
Why can a decaying learning rate slow adaptation to a genuinely new pattern late in a stream?
Hints
- Compare update sizes early and late.
- Convergence and tracking a moving target differ.
Answer
This answer is part of mentoring
Advanced questions come with a full breakdown and a mock interview with a working quant researcher. The first call is free.
Source and licence: scikit-learn 1.7.2 — scaling strategies · BSD-3-Clause · answer: Quantoros editorial