Theory Projected_Gradient_Descent
theory Projected_Gradient_Descent
imports Main_Results
begin
section ‹Projected Gradient Descent›
text ‹
This is the umbrella theory of the entry.
It imports the complete public development for smooth convex first-order
optimization through @{text ‹Main_Results›}. The imported material includes reusable
interfaces for gradients, convex first-order reasoning, smooth upper bounds,
abstract descent arguments, projection geometry, projected-gradient descent,
projected-gradient mappings, residual convergence certificates, strong
convexity, linear-rate refinements, Lipschitz-smoothness bridges, and
quadratic examples.
The recommended public theorem surface is collected in @{text ‹Main_Results›}.
Downstream developments should normally import @{text ‹Main_Results›} when they want to
reuse the library, rather than depending on the internal organization of the
individual proof files.
›
end