We introduce and analyze a natural geometric version of Renegar's condition number $\mathcal{R}$ for the homogeneous convex feasibility problem associated with a regular cone $C\subseteq\mathbb{R}^n$. Let ${\rm Gr}_{n,m}$ denote the Grassmann manifold of $m$-dimensional linear subspaces of $\mathbb{R}^n$, and consider the projection distance $d_p(W_1,W_2) := \| \Pi_{W_1} - \Pi_{W_2}\|$ (spectral norm) between $W_1,W_2\in {\rm Gr}_{n,m}$, where $\Pi_{W_i}$ denotes the orthogonal projection onto $W_i$. We call $\mathscr{C}(W) := \max\{ d_p(W,W')^{-1} \mid W' \in \Sigma_m\}$ the Grassmann condition number of $W\in{\rm Gr}_{n,m}$, where the set of ill-posed instances $\Sigma_m\subset{\rm Gr}_{n,m}$ is defined as the set of linear subspaces touching $C$. We show that if $W ={\rm im}(A^T)$ for a matrix $A\in\mathbb{R}^{m\times n}$, then $\mathscr{C}(W) \le \mathcal{R}(A) \le \mathscr{C}(W)\, \kappa(A)$, where $\kappa(A) =\|A\| \|A^\dagger\|$ denotes the matrix condition number. This extends work by Belloni and Freund in [Math. Program. Ser. A, 119 (2009), pp. 95--107]. Furthermore, we show that $\mathscr{C}(W)$ can also be characterized in terms of the Riemannian distance metric on ${\rm Gr}_{n,m}$. This differential geometric characterization of $\mathscr{C}(W)$ is the starting point of the sequel [D. Amelunxen and P. Bürgisser, Probabilistic analysis of the Grassman condition number, preprint, http:arxiv.org/abs/1112.2603, 2011] to this paper, where the first probabilistic analysis of Renegar's condition number for an arbitrary regular cone $C$ is achieved.
Read full abstract