scalanlp/breeze

Breeze LBFGS Preconditionner

Open

#271 opened on Jul 1, 2014

View on GitHub
 (0 comments) (0 reactions) (0 assignees)Scala (3,453 stars) (690 forks)batch import
enhancementhelp wantedproject

Description

When I seed the LBFGS with a vector very close to the optimal solution, the first pass move away from the optimal, then returns after few passes. The issue is about the implementing custom preconditioner (like the standard Jacobi preconditioner)

Contributor guide

Tech stack
None
Domain
backenddata
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
3-5 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
Basic understanding of LBFGS algorithmFamiliarity with ScalaKnowledge of numerical optimizationExperience with Breeze library
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
20
Research direction
Examine the existing LBFGS implementation in Breeze (likely in the breeze/src/main/scala/breeze/optimize/ directory). Understand the current preconditioner interface and how LBFGS uses it. Look for any related issues or PRs about preconditioners. Implement a Jacobi preconditioner as a starting point, following the pattern of existing optimization algorithms in the library. Consider writing tests and documentation for the new feature.
Breeze LBFGS Preconditionner · scalanlp/breeze#271 | Good First Issue