Good first issueNew feature
Repository metrics
- Stars
- (4,574 stars)
- PR merge metrics
- (平均マージ 47d 6h) (30d で 46 merged PRs)
説明
It seems RLS is a way to reach the same solution as batch linear regression. It's probably more expensive than SGD, but is still interesting.
Resources
- https://www.wikiwand.com/en/Recursive_least_squares_filter
- https://people.eecs.berkeley.edu/~jiantao/225a2020spring/scribe/EECS225A_Lecture_15.pdf
- http://pfister.ee.duke.edu/courses/ece586/ex_proj_2008.pdf
- Deterministic Online Classification: Non-iteratively Reweighted Recursive Least-Squares for Binary Class Rebalancing