Composite Generalized Smooth Minimization
Descriptions may be stale. Consult the linked Lean sources.
Composite Generalized Smooth Minimization
We consider minimizing
The optimization problem is
Our assumptions:
Assumption 1 (Convexity). The set
Assumption 2 (Generalized smoothness). The function
Assumption 3 (Exact proximal access). There is a map
Assumption 4 (Initialization and minimizer bounds).
The given starting point satisfies
The oracle returns
After the allotted calls, the output must satisfy
Results
Minimize
| Result | Oracle complexity | Novelty | Limitations |
|---|---|---|---|
| (1) CompositeGeneralizedAccelerated
(Ours) |
|||
| (2) (Tyurin 2026),
Thm. 4.3. Special case: |
(1): Scope extends
acceleration to constrained/composite objectives with |
(1): Larger logarithmic
term can occur when |
|
| (3) AC-FGM (Li and Lan 2024),
Cor. 1 and local-smoothness remark. Finite-dimensional, closed |
(1): No function values;
explicit accelerated composite bound in setup parameters. Rates are not
comparable through an unspecified trajectory curvature |
(1): Needs |
Abbr: Each gradient, value or prox evaluation counts
as one call. A prox query minimizes