What's new in LIBLINEAR 2.47
Jul 9, 2023
New in LIBLINEAR 2.46 (Feb 17, 2023)
- We improve the training speed of the one-class SVM solver.
New in LIBLINEAR 2.45 (Aug 11, 2022)
- Some minor bugs are fixed.
New in LIBLINEAR 2.44 (Mar 17, 2022)
- Some minor bugs are fixed.
New in LIBLINEAR 2.41 (Jul 29, 2020)
- A new solver: dual coordinate descent method for linear one-class SVM; see the paper
- The Newton solver is updated to have faster training speed; see the release note
- A new option -R to allow users not to regularize bias (when -B 1 is used)
New in LIBLINEAR 2.21 (Oct 5, 2018)
- Since 2.20 we have continued to improve the conjugate gradient method in primal-based Newton solvers by applying a preconditioning technique. The convergence can be significantly improved on some problems (see the paper).
- In 2.21, some python interface files are re-organirzed (following the changes in LIBSVM 3.23)
New in LIBLINEAR 2.20 (Apr 12, 2018)
- We have improved the conjugate gradient method in primal-based Newton solvers by applying a preconditioning technique.
- The convergence can be significantly improved on some problems
New in LIBLINEAR 1.96 (Nov 15, 2014)
- It conducts some minor fixes. Also note that we now provide 64-bit windows binary exe files.