Hessian-Stable Skew Monotone Equations
Descriptions may be stale. Consult the linked Lean sources.
Hessian-Stable Skew Monotone Equations
We seek a zero of a differentiable map
The underlying equation is
Our assumptions:
Assumption 1 (Skew decomposition and multiplicative
stability). The map
Assumption 2 (Nearby zero and local Jacobian bound).
There exists
Queries to either oracle may be made at any point of
Results
On a finite-dimensional real inner-product space,
| Result | Oracle complexity | Novelty | Limitations |
|---|---|---|---|
| (1) HessianStableSkewNewton
(Ours) |
|||
| (2) AEP (Jiang and Yoon 2026), Thms. 4.4, 4.6 | (1):
Polynomial-to-polylogarithmic accuracy gain on the overlap as |
(1) requires constant
skew and Hessian stability; can cost more at moderate accuracy and large
|
|
| (3) EAG (Yoon and Ryu 2021), Thm. 1, §2.3 | (1): Lower cost on the
overlap when |
(1) needs |
|
| (4) MS-BACON (Carmon et al. 2020), Cor. 12 | (1): Scope extends
gradient fields to constant skew coupling. Matches the |
||
| (5) Regularized Newton (Doikov 2023), Thm. 3.3 | (1): Scope extends
minimization to constant skew coupling. Rates are not directly
comparable using |
Abbr: