Simpler Grassmannian optimization
Zehua Lai,Lek-Heng Lim,Ke Ye
DOI: https://doi.org/10.48550/arXiv.2009.13502
2020-09-29
Abstract:There are two widely used models for the Grassmannian $\operatorname{Gr}(k,n)$, as the set of equivalence classes of orthogonal matrices $\operatorname{O}(n)/(\operatorname{O}(k) \times \operatorname{O}(n-k))$, and as the set of trace-$k$ projection matrices $\{P \in \mathbb{R}^{n \times n} : P^{\mathsf{T}} = P = P^2,\; \operatorname{tr}(P) = k\}$. The former, standard in manifold optimization, has the advantage of giving numerically stable algorithms but the disadvantage of having to work with equivalence classes of matrices. The latter, widely used in coding theory and probability, has the advantage of using actual matrices (as opposed to equivalence classes) but working with projection matrices is numerically unstable. We present an alternative that has both advantages and suffers from neither of the disadvantages; by representing $k$-dimensional subspaces as symmetric orthogonal matrices of trace $2k-n$, we obtain \[ \operatorname{Gr}(k,n) \cong \{Q \in \operatorname{O}(n) : Q^{\mathsf{T}} = Q, \; \operatorname{tr}(Q) = 2k -n\}. \] As with the other two models, we show that differential geometric objects and operations -- tangent vector, metric, normal vector, exponential map, geodesic, parallel transport, gradient, Hessian, etc -- have closed-form analytic expressions that are computable with standard numerical linear algebra. In the proposed model, these expressions are considerably simpler, a result of representing $\operatorname{Gr}(k,n)$ as a linear section of a compact matrix Lie group $\operatorname{O}(n)$, and can be computed with at most one QR decomposition and one exponential of a special skew-symmetric matrix that takes only $O(nk(n-k))$ time. In particular, we completely avoid eigen- and singular value decompositions in our steepest descent, conjugate gradient, quasi-Newton, and Newton methods for the Grassmannian.
Optimization and Control