Ask HN: Is there a taxonomy of machine learning types?

Besides classification and regression, and the unsupervised methods for principle components, clustering and frequent item-sets, what tools are there in the ML toolkit and what kinds of problems are amenable to their use?