Book Memo: “Mathematical Theories of Machine Learning”

Theory and Applications
This information analysis mathematical theories of machine finding out. The first half of the information explores the optimality and adaptivity of choosing step sizes of gradient descent for escaping strict saddle components in non-convex optimization points. In the second half, the authors counsel algorithms to go looking out native minima in nonconvex optimization and to accumulate worldwide minima in a degree from the Newton Second Law with out friction. In the third half, the authors study the problem of subspace clustering with noisy and missing information, which is a matter well-motivated by smart capabilities information subject to stochastic Gaussian noise and/or incomplete information with uniformly missing entries. In the ultimate half, the authors introduce an novel VAR model with Elastic-Net regularization and its equal Bayesian model allowing for every a gradual sparsity and a gaggle selection.