New solver for MKL-SVM with 0/1 loss function.
problem Optimizing MKL-SVM with -loss function.
method Developed an ADMM solver for nonconvex, nonsmooth optimization.
result Shows promise in simple numerical experiment.
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
New solver for MKL-SVM with 0/1 loss function.
Paper introduces MKL--SVM for SVM with loss.