Regularization of limited memory quasi-Newton methods for large-scale nonconvex minimization.
Christian KanzowDaniel SteckPublished in: Math. Program. Comput. (2023)
Keyphrases
- limited memory
- quasi newton method
- image restoration and reconstruction
- objective function
- regularization term
- memory space
- sliding window
- data streams
- influence diagrams
- half quadratic
- minimization problems
- real time
- stream mining
- markov random field
- convex optimization
- past observations
- optimization problems
- convex functions
- special case
- knowledge base
- bayesian networks
- norm minimization
- nonlinear programming
- lower bound
- dynamic programming
- total variation