keystoneml.nodes.learning
Run Limited-memory BFGS (L-BFGS) in parallel.
Run Limited-memory BFGS (L-BFGS) in parallel. Averaging the subgradients over different partitions is performed using one standard spark map-reduce in each iteration.