Online Stochastic Gradient Descent with Arbitrary Initialization Solves Non-smooth, Non-convex Phase Retrieval
This provides a theoretical guarantee for a simpler approach to phase retrieval, which is important for signal processing and imaging applications, though it is incremental as it builds on existing methods.
The paper proves that online stochastic gradient descent (SGD) with constant step size converges from arbitrary initializations for the non-smooth, non-convex phase retrieval problem using the amplitude squared loss, eliminating the need for a spectral initialization step and showing equivalence to the randomized Kaczmarz algorithm.
In recent literature, a general two step procedure has been formulated for solving the problem of phase retrieval. First, a spectral technique is used to obtain a constant-error initial estimate, following which, the estimate is refined to arbitrary precision by first-order optimization of a non-convex loss function. Numerical experiments, however, seem to suggest that simply running the iterative schemes from a random initialization may also lead to convergence, albeit at the cost of slightly higher sample complexity. In this paper, we prove that, in fact, constant step size online stochastic gradient descent (SGD) converges from arbitrary initializations for the non-smooth, non-convex amplitude squared loss objective. In this setting, online SGD is also equivalent to the randomized Kaczmarz algorithm from numerical analysis. Our analysis can easily be generalized to other single index models. It also makes use of new ideas from stochastic process theory, including the notion of a summary state space, which we believe will be of use for the broader field of non-convex optimization.