Optimization All non-trivial statistical estimations, computationally speaking, result in optimization problems; increasingly powerful models generally lead to more difficult optimizations. fastlab Big Ideas People Stuff Isee elsewhere Scalable Semidefinite Programming We showed an approach for scalable learning based on semidefinite programming based on convex relaxations, in the context of manifold learning. [see full entry here] Iin progress Faster Optimization for Kernel Machines We are developing faster optimization methods for both batch and online training of a general class of regularized learning objectives, which includes support vector machines. We will also show a framework which includes recent online algorithms as special cases while also illuminating new variants. Faster Mixed Integer Programming for Kernel Machines In the context of L0 support vector machines, we have developed more efficient mixed integer programming formulations based on recent optimization techniques.