Release of MLSVM 1.1, Multilevel Support Vector Machines

[Ilya Safro]

We are pleased to announce the release of MLSVM 1.1, a library of fast multilevel algorithms for training nonlinear support vector machine models on large-scale datasets. The library is developed as an extension of PETSc to support, among other applications, the analysis of datasets in scientific computing.

Version 1.1 includes several accelerations and bug fixes.

Highlights:
- The best quality/performance trade-off is achieved with algebraic multigrid coarsening
- Tested on academic, industrial, and healthcare datasets
- Generates multiple models for each training
- Effective on imbalanced datasets

Download MLSVM at https://github.com/esadr/mlsvm

Corresponding paper: Sadrfaridpour, Razzaghi and Safro "Engineering multilevel support vector machines", 2017, https://arxiv.org/pdf/1707.07657.pdf