Online Learning as Stochastic Approximation of Regularization Paths: Optimality and Almost-Sure Convergence | IEEE Journals & Magazine | IEEE Xplore