Name
Abstract | ||
|---|---|---|
Author(s): Anandkumar, A; Deng, Y; Ge, R; Mobahi, H | Abstract: Developing efficient and guaranteed nonconvex algorithms has been an important challenge in modern machine learning. Algorithms with good empirical performance such as stochastic gradient descent often lack theoretical guarantees. In this paper, we analyze the class of homotopy or continuation methods for global optimization of nonconvex functions. These methods start from an objective function that is efficient to optimize (e.g. convex), and progressively modify it to obtain the required objective, and the solutions are passed along the homotopy path. For the challenging problem of tensor PCA, we prove global convergence of the homotopy method in the high regime. The signal-to-noise requirement for our algorithm is tight in the sense that it matches the recovery guarantee for the best degree-4 sum-of-squares algorithm. In addition, we prove a phase transition along the homotopy path for tensor PCA. This allows to simplify the homotopy method to a local search algorithm, viz., tensor power iterations, with a specific initialization and a noise injection procedure, while retaining the theoretical guarantees. |
Year | Venue | Field |
|---|---|---|
2017 | COLT | Convergence (routing),Mathematical optimization,Stochastic gradient descent,Tensor,Global optimization,Artificial intelligence,Local search (optimization),Homotopy,Initialization,Homotopy analysis method,Machine learning,Mathematics |
DocType | Citations | PageRank |
Conference | 4 | 0.43 |
References | Authors | |
12 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Animashree Anandkumar | 1 | 1629 | 116.30 |
| Yuan Deng | 2 | 25 | 11.91 |
| Rong Ge | 3 | 1626 | 76.01 |
| Hossein Mobahi | 4 | 382 | 26.89 |