An ADMM Solver for the MKL-$L_{0/1}$-SVM
This is an incremental improvement for machine learning researchers working on kernel methods and optimization.
The authors tackled the problem of multiple kernel learning for support vector machines with a nonconvex (0,1)-loss function by developing an ADMM solver, and a simple numerical experiment on synthetic data suggested it could be promising.
We formulate the Multiple Kernel Learning (abbreviated as MKL) problem for the support vector machine with the infamous $(0,1)$-loss function. Some first-order optimality conditions are given and then exploited to develop a fast ADMM solver for the nonconvex and nonsmooth optimization problem. A simple numerical experiment on synthetic planar data shows that our MKL-$L_{0/1}$-SVM framework could be promising.