GNBO
GNBO is a generalized Newton bilevel optimization framework that unifies Newton-type and quasi-Newton-type bilevel methods under a shared curvature structure, and is used to analyze convergence and design new algorithms.
Timeline (last 90 days)
A recent paper proposes two Newton methods that go beyond lower-level strong convexity: Newton GALET and Newton BAMM, and introduces a unified framework, GNBO.
The GNBO framework captures the shared curvature structure of the proposed methods and several existing Newton and quasi-Newton bilevel methods.
A unique insight from a single source points out that the GNBO framework may have inspired the design of subsequent quasi-Newton methods.
All Facts (3)
A recent paper proposes two Newton methods that go beyond lower-level strong convexity: Newton GALET and Newton BAMM, and introduces a unified framework, GNBO.
50%UnverifiedThe GNBO framework captures the shared curvature structure of the proposed methods and several existing Newton and quasi-Newton bilevel methods.
50%UnverifiedA unique insight from a single source points out that the GNBO framework may have inspired the design of subsequent quasi-Newton methods.
50%