Abstract:We present a notion of bilinear stability, which is to numerical stability what bilinear complexity is to time complexity. In bilinear complexity, an algorithm for evaluating a bilinear operator $\beta : \mathbb{U} \times \mathbb{V} \to \mathbb{W}$ is a decomposition $\beta = \varphi_1 \otimes \psi_1 \otimes w_1 + \dots + \varphi_r \otimes \psi_r \otimes w_r $; the number of terms $r$ captures the speed of the algorithm; and its smallest possible value, i.e., the tensor rank of $\beta$, quantifies the speed of a fastest algorithm. Bilinear stability introduces norms to the mix: The growth factor of the algorithm $\lVert \varphi_1 \rVert_* \lVert \psi_1 \rVert_* \lVert w_1 \rVert + \dots + \lVert \varphi_r \rVert_* \lVert \psi_r \rVert_* \lVert w_r \rVert$ captures the accuracy of the algorithm; and its smallest possible value, i.e., the tensor nuclear norm of $\beta$, quantifies the accuracy of a stablest algorithm. To substantiate this notion, we establish a bound for the forward error in terms of the growth factor and present numerical evidence comparing various fast algorithms for matrix and complex multiplications, showing that larger growth factors correlate with less accurate results. Compared to similar studies of numerical stability, bilinear stability is more general, applying to any bilinear operators and not just matrix or complex multiplications; is more simplistic, bounding forward error in terms of a single (growth) factor; and is truly tensorial like bilinear complexity, invariant under any orthogonal change of coordinates. As an aside, we study a new algorithm for computing complex multiplication in terms of real, much like Gauss's, but is optimally fast and stable in that it attains both tensor rank and nuclear norm.

Numerical solution to a linear equation with tensor product structure.

Numerical stability and tensor nuclear norm

Algebraic Lyapunov and Stein Stability Results for Tensors

Developing iterative algorithms to solve Sylvester tensor equations

A nested divide-and-conquer method for tensor Sylvester equations with positive definite hierarchically semiseparable coefficients

Solving differential Riccati equations: A nonlinear space-time method using tensor trains

Solving Multi-linear Systems with ℳ -Tensors

Solutions of linear and non-linear partial differential equations by means of tensor product theory of Banach space

Tensor Train Decomposition for Solving Large-Scale Linear Equations.

Iterative Solution of A System of Nonlinear Algebraic Equations F(X)=0, Using (X) over Dot = Lambda[Alpha R Plus Beta P] or (X) over Dot = Lambda[Alpha F Plus Beta P] R is A Normal to A Hyper-Surface Function of F, P Normal to R, and P Normal to F

Tensor Krylov subspace methods via the T-product for large Sylvester tensor equations

MODIFIED SPLITTING METHODS FOR SOLVING NON-HOMOGENOUS MULTI-LINEAR EQUATIONS WITH M-TENSORS

Preconditioned low-rank Riemannian optimization for linear systems with tensor product structure

Several efficient iterative algorithms for solving nonlinear tensor equation with Einstein product

Numerical algorithms for the fast and reliable solution of periodic tridiagonal Toeplitz linear systems

Randomized Coordinate Descent Method for Inconsistent Tensor Linear Systems with T-Product.

A tensor-based approach to solving linear systems involving Kronecker sum of matrices

Numerical Solution to Linear Matrix Equation by Finite Steps Iteration

Randomized extended average block Kaczmarz method for inconsistent tensor equations under t-product

Tensor Methods for Solving Symmetric ℳ -tensor Systems

Solving Multi-linear Systems with M-Tensors.