Abstract:In this paper, we consider the geometric landscape connection of the widely studied manifold and factorization formulations in low-rank positive semidefinite (PSD) and general matrix optimization. We establish an equivalence on the set of first-order stationary points (FOSPs) and second-order stationary points (SOSPs) between the manifold and the factorization formulations. We further give a sandwich inequality on the spectrum of Riemannian and Euclidean Hessians at FOSPs, which can be used to transfer more ge… Show more
“…Although the convex relaxation is usually guaranteed to recover the exact ground truth with almost the optimal sample complexity, the associated algorithms operate in the space of matrix variables and, thus, are computationally inefficient for largescale problems [60]. Similar issues are observed for algorithms based on the Singular Value Projection [31] and Riemannian optimization algorithms [53,54,30,2,37]. The analysis of the convex relaxation approach in the noisy case is recently conducted by bridging the convex and the nonconvex approaches [21,22].…”
In this work, we develop a new complexity metric for an important class of low-rank matrix optimization problems, where the metric aims to quantify the complexity of the nonconvex optimization landscape of each problem and the success of local search methods in solving the problem. The existing literature has focused on two complexity measures. The RIP constant is commonly used to characterize the complexity of matrix sensing problems. On the other hand, the sampling rate and the incoherence are used when analyzing matrix completion problems. The proposed complexity metric has the potential to unify these two notions and also applies to a much larger class of problems. To mathematically study the properties of this metric, we focus on the rank-1 generalized matrix completion problem and illustrate the usefulness of the new complexity metric from three aspects. First, we show that instances with the RIP condition have a small complexity. Similarly, if the instance obeys the We note that a similar complexity metric based on a special case of instances in Section 3.3 was proposed in our conference paper [55]. However, the complexity metric in this work has a different form and is proved to work on a broader set of applications. In addition, we prove several theoretical properties of the metric in this work, which are not included in [55].
“…Although the convex relaxation is usually guaranteed to recover the exact ground truth with almost the optimal sample complexity, the associated algorithms operate in the space of matrix variables and, thus, are computationally inefficient for largescale problems [60]. Similar issues are observed for algorithms based on the Singular Value Projection [31] and Riemannian optimization algorithms [53,54,30,2,37]. The analysis of the convex relaxation approach in the noisy case is recently conducted by bridging the convex and the nonconvex approaches [21,22].…”
In this work, we develop a new complexity metric for an important class of low-rank matrix optimization problems, where the metric aims to quantify the complexity of the nonconvex optimization landscape of each problem and the success of local search methods in solving the problem. The existing literature has focused on two complexity measures. The RIP constant is commonly used to characterize the complexity of matrix sensing problems. On the other hand, the sampling rate and the incoherence are used when analyzing matrix completion problems. The proposed complexity metric has the potential to unify these two notions and also applies to a much larger class of problems. To mathematically study the properties of this metric, we focus on the rank-1 generalized matrix completion problem and illustrate the usefulness of the new complexity metric from three aspects. First, we show that instances with the RIP condition have a small complexity. Similarly, if the instance obeys the We note that a similar complexity metric based on a special case of instances in Section 3.3 was proposed in our conference paper [55]. However, the complexity metric in this work has a different form and is proved to work on a broader set of applications. In addition, we prove several theoretical properties of the metric in this work, which are not included in [55].
“…They showed while the sets of FOSPs under the factorization formulation can be larger, the sets of SOSPs are contained in the set of fixed points of the PGD with a small stepsize. Another related work is [LLZ21], where they studied the landscape connections of the factorization formulation and the Riemannian formulation with embedded geometry for both (1) and (2) and identified the close connections between the landscapes under these two formulations, e.g., the set of FOSPs and SOSPs under these two formulations are exactly the same when constraining to rank-r matrices. In addition, under the quotient geometry, a general equivalence relation on the sets of FOSPs and SOSPs of objectives on the total space and the quotient space is given in [Bou20,Section 9.11].…”
Section: Related Literaturementioning
confidence: 99%
“…As we have discussed in the Introduction, another popular approach for handling the rank constraint in (1) or (2) is via factorizing X into YY J or LR J and then treat the new problem as unconstrained optimization in the Euclidean space. In the recent work [LLZ21], they showed for both (1) and (2) the geometric landscapes under the factorization and embedded submanifold formulations are almost equivalent. By combining their results and the results in this paper, we also have a geometric landscape connection of (1) and (2) under the factorization and quotient manifold formulations.…”
Section: Geometric Connections Of Embedded and Quotient Geometries In...mentioning
In this paper, we propose a general procedure for establishing the landscape connections of a Riemannian optimization problem under the embedded and quotient geometries. By applying the general procedure to the fixed-rank positive semidefinite (PSD) and general matrix optimization, we establish an exact Riemannian gradient connection under two geometries at every point on the manifold and sandwich inequalities between the spectra of Riemannian Hessians at Riemannian first-order stationary points (FOSPs). These results immediately imply an equivalence on the sets of Riemannian FOSPs, Riemannian second-order stationary points (SOSPs) and strict saddles of fixed-rank matrix optimization under the embedded and the quotient geometries. To the best of our knowledge, this is the first geometric landscape connection between the embedded and the quotient geometries for fixed-rank matrix optimization and it provides a concrete example on how these two geometries are connected in Riemannian optimization. In addition, the effects of the Riemannian metric and quotient structure on the landscape connection are discussed. We also observe an algorithmic connection for fixed-rank matrix optimization under two geometries with some specific Riemannian metrics. A number of novel ideas and technical ingredients including a unified treatment for different Riemannian metrics and new horizontal space representations under quotient geometries are developed to obtain our results. The results in this paper deepen our understanding of geometric connections of Riemannian optimization under different Riemannian geometries and provide a few new theoretical insights to unanswered questions in the literature.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.