A Semismooth-Newton's-Method-Based Linearization and Approximation Approach for Kernel Support Vector Machines.
Chen JiangQingna LiPublished in: CoRR (2020)
Keyphrases
- synthetic data
- significant improvement
- theoretical analysis
- support vector machine svm
- pairwise
- dynamic programming
- experimental evaluation
- approximation methods
- computational cost
- support vector machine
- detection method
- closed form
- polygonal approximation
- data mining
- sensitivity analysis
- optimization method
- classification method
- segmentation method
- clustering method
- high accuracy