Twice differentiable spectral functions
The basic elements of the paper are \(S^{n}\) as the Euclidean space of all \(n\times n\) symmetric matrices with inner product \(<A,B>=\)tr\((A\cdot B)\), the vector \(\lambda (A)=(\lambda_{1}(A),\lambda_{2}(A),\dots\lambda{n}(A))\) of its eigenvalues ordered in nonincreasing order and a symmetric function \(f:\mathbb{R}^{n}\mapsto \mathbb{R}\). Firstly, the authors prove some results on \(S^{n}\) and for the symmetric function (\(f(x)=f(P(x))\) for any permutation matrix \(P\) on \(x\)).NEWLINENEWLINEThe main result gives the condition for a symmetric function to be twice differentiable. So, the symmetric function \(f\) is \(C^{2}\) at the point \(\mu\) iff \(f\circ \lambda\) is \(C^{2}\) at the point Diag\(\mu\). For a symmetric matrix \(A\) the symmetric function \(f\) is \(C^{2}\) at the point \(\lambda (A)\) iff the spectral function \(f\circ \lambda\) is \(C^{2}\) at the matrix \(A\). In the both theorems the expression of the Hessian is given.
- Isotropic functions revisited
- A unified approach to operator monotone functions
- Smooth convex approximation to the maximum eigenvalue function
- On the connectedness of spectral sets and irreducibility of spectral cones in Euclidean Jordan algebras
- Speckle reduction in matrix-log domain for synthetic aperture radar imaging
- The degrees of freedom of partly smooth regularizers
- Spectral operators of matrices
- Clarke generalized Jacobian of the projection onto the cone of positive semidefinite matrices
- An introduction to a class of matrix cone programming
- On the uniqueness of solutions of spectral equations
- A spectral quadratic-SDP method with applications to fixed-order \(H_2\) and \(H_\infty\) synthesis
- Scalar curvature and deformations of complex structures
- Spectral (isotropic) manifolds and their dimension
- Twice Fréchet differentiability of some functions on \(\mathbb{R}^n\)
- Spectral self-concordant functions in the space of two-by-two symmetric matrices
- Low-rank matrix iteration using polynomial-filtered subspace extraction
- A smoothing majorization method for l^2_2-l^p_p matrix minimization
- On the central paths and Cauchy trajectories in semidefinite programming
- scientific article; zbMATH DE number 3910560 (Why is no real title available?)
- Variational analysis of spectral functions simplified
- scientific article; zbMATH DE number 7033669 (Why is no real title available?)
- Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising
- A unified approach to spectral and isotropic functions
- On the analyticity of k-isotropic functions
- Derivatives of Spectral Functions
- Second-order conditions for existence of augmented Lagrange multipliers for eigenvalue composite optimization problems
- Derivatives of compound matrix valued functions
- Spectral sets and functions on Euclidean Jordan algebras
- An approximation scheme for distributionally robust PDE-constrained optimization
- Some New Methods for Generating Convex Functions
- Spectral operators of matrices: semismoothness and characterizations of the generalized Jacobian
- Generalized Hadamard Product and the Derivatives of Spectral Functions
- Analysis of Symmetric Matrix Valued Functions
- On the analysis of the discretized Kohn-Sham density functional theory
- On twice differentiable functions
- Quadratic expansions of spectral functions
- Truncated nonsmooth Newton multigrid for phase-field brittle-fracture problems, with analysis
- A semismooth Newton based dual proximal point algorithm for maximum eigenvalue problem
- Spectrally constrained optimization
- On a minimization problem of the maximum generalized eigenvalue: properties and algorithms
- Variational analysis of orthogonally invariant norm cones of symmetric matrices
- Nonsingularity of Karush-Kuhn-Tucker system in a class of linear matrix conic optimization problems
- Entropic covariance models
- A feasible smoothing accelerated projected gradient method for nonsmooth convex optimization
- Smooth optimization using global and local low-rank regularizers
- Second-order directional derivatives of spectral functions
- Properties of the augmented Lagrangian in nonlinear semidefinite optimization
- Convexity and differentiability properties of spectral functions and spectral mappings on Euclidean Jordan algebras
- Smoothing technique and its applications in semidefinite optimization
- The higher-order derivatives of spectral functions
- Analysis of symmetric matrix valued functions. I
- Sensitivity analysis of the strain criterion for multidimensional scaling
This page was built for publication: Twice differentiable spectral functions
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2784352)
