betaFeedback

WorksMathematics

Anderson Acceleration of SCF Iteration and Newton's Method for Solving Nonlinear Eigenvector Problems

Abstract

The self-consistent field (SCF) iteration is a gateway method for solving nonlinear eigenvector problems, also known as nonlinear eigenvalue problems with eigenvector dependency (NEPv). However, the convergence of SCF can be slow in practice. In this paper, we study two local acceleration techniques for SCF applied to unitarily invariant NEPv: Anderson acceleration and Newton acceleration. A central difficulty in developing these techniques is the non-uniqueness of the eigenvector basis, which complicates direct application of standard Anderson acceleration and introduces singularity into the correction equation of the standard Newton formulation. For Anderson acceleration, we present three formulations based on unitarily invariant parameterization, basis alignment, and Grassmannian geometry, and establish regularity properties of the associated SCF maps. For Newton acceleration, we introduce basis-aligned updates and derive a correction equation whose operator is nonsingular under suitable assumptions, and prove local quadratic convergence. Numerical experiments on NEPv arising from computational physics and data analysis confirm the theoretical findings and demonstrate the effectiveness of the proposed acceleration techniques.

Introduction

We consider the following nonlinear eigenvalue problem with eigenvector dependency (NEPv): find a matrix X∈Cn,kX \in\mathbb{C}^{n,k} with orthonormal columns (i.e., XHX=IkX^{H}X=I_k) and a square matrix Λ∈Ck,k\Lambda\in\mathbb{C}^{k,k} satisfying

H(X)X=XΛ,(1)H(X)X=X\Lambda, \tag*{(1)}

where H:Cn,k→Cn,nH:\mathbb{C}^{n,k}\to\mathbb{C}^{n,n} is a given Hermitian matrix-valued function. Here, the eigenvalues of Λ≡XHH(X)X\Lambda\equiv X^{H}H(X)X correspond to a subset of kk eigenvalues of H(X)H(X). We assume that they are the kk smallest throughout the paper. The largest or those at other prescribed positions can be treated analogously. Typically, k≪nk \ll n. In addition, we assume that HH is right unitarily invariant, i.e.,

H(X)=H(XQ),for all Q∈Ck,k with QHQ=Ik.(2)H(X)=H(XQ), \qquad\text{for all } Q \in\mathbb{C}^{k,k} \text{ with } Q^{H}Q=I_k. \tag*{(2)}

This implies that any solution (X,Λ)(X,\Lambda) would give rise to an equivalent class of solution (XQ,QHΛQ)(XQ,Q^{H}\Lambda Q). Therefore, a solution XX is intrinsically the subspace range⁡(X)\operatorname{range}(X).

NEPv (1) arises in a wide range of applications. They have long played important roles in quantum physics and chemistry through the Hartree–Fock and Kohn–Sham equations [10, 16, 19, 21, 39, 55]. More recently, they have attracted growing interest in machine learning, with applications including sparse PCA, trace ratio optimization, and Rayleigh quotient optimization; see, e.g., [9, 14, 53, 70, 71, 75, 76, 77]. The widespread use of NEPv stems from their fundamental role as first-order optimality conditions for optimization with orthogonality constraints, and the invariance condition (2) reflects that the solution sought is a subspace rather than a particular basis in those applications; see, e.g., [48].

A standard approach to solving (1) is the self-consistent field (SCF) iteration. Starting from an initial matrix X0X_0 with orthonormal columns, SCF repeatedly solves the linear eigenvalue problem

H(Xi)Xi+1=Xi+1Λi+1,i=0,1,2,…,(3)H(X_i)X_{i+1}=X_{i+1}\Lambda_{i+1}, \qquad i=0,1,2,\ldots, \tag*{(3)}

where the columns of Xi+1X_{i+1} are orthonormal eigenvectors corresponding to the kk smallest eigenvalues of H(Xi)H(X_i), and Λi+1\Lambda_{i+1} is Hermitian but need not be diagonal. We note that the iterate matrix Xi+1X_{i+1} is generally not unique, with different choices related by an orthonormal rotation Xi+1QX_{i+1}Q, provided that the kk smallest eigenvalues are separated from the rest of the spectrum. By the unitary invariance (2), these choices give the same H(Xi+1)H(X_{i+1}) and thus do not affect the subsequent iterations. Hence, the SCF (3) is an iteration on subspaces {range⁡(Xi)}i=0∞\{\operatorname{range}(X_i)\}_{i=0}^{\infty}, with each XiX_i serving as a representative orthonormal basis.

The SCF iteration was originally developed for electronic structure calculations [33, 60]. Under suitable assumptions, local convergence can be established, typically at a linear rate [7, 17, 51, 68, 74]. In practice, slow convergence or failure to converge may occur and has motivated various techniques to stabilize and accelerate SCF, such as damping, level shifting, and mixing of iterates [20, 45, 58, 59, 64, 66]. However, many of these techniques have been developed for specific computational and physical applications, such as density functional theory, where they exploit particular functional structures of H(X)H(X), and they do not readily apply to general NEPv (1) in a computationally efficient way.

The purpose of this paper is to investigate acceleration techniques of the SCF iteration (3) for unitarily invariant NEPv (1) in its most general form. We focus on two types of acceleration: Anderson acceleration (AA), which combines recent iterates, and Newton acceleration, which uses derivative information of H(X)H(X) to achieve superlinear acceleration.

AA is a widely used technique for accelerating fixed-point iterations [5]. The basic idea is to form a refined approximation through a linear combination (or mixing) of recent iterates, using suitably chosen coefficients; see, e.g., [6, 26, 27, 67, 69] for details. Applying AA to the SCF (3) requires extra care because the iterates XiX_i are not uniquely determined, and directly mixing them depends on the arbitrary unitary rotations XiQX_iQ. In electronic structure calculations, AA is closely related to Pulay mixing, also known as direct inversion in the iterative subspace (DIIS) [58, 59]. For these specialized NEPv, the SCF can be expressed in terms of quantities such as the charge density diag⁡(XXH)\operatorname{diag}(XX^{H}), density matrix XXHXX^{H}, or Fock matrix H(X)H(X), which are independent of the chosen basis; see, e.g., [30, 36, 46, 59]. For example, for the Kohn–Sham NEPv that arises in density functional theory, where H(X)H(X) is fully determined by the charge density ρ(X)=diag⁡(XXH)∈Rn\rho(X)=\operatorname{diag}(XX^{H}) \in\mathbb{R}^{n}, the SCF can be recast as a fixed-point iteration ρi+1=g(ρi)\rho_{i+1}=g(\rho_i) in charge densities, and AA can then be applied directly to this iteration; see, e.g., [63]. However, for general NEPv (1) such a compact density-based representation of H(X)H(X) may not be readily available, while mixing the full matrices H(Xi)∈Cn,nH(X_i) \in\mathbb{C}^{n,n} can be costly for large nn. This calls for an in-depth investigation of AA in these settings, particularly into how to work economically with tall-skinny basis matrices XiX_i.

On the other hand, Newton’s method provides another means of locally accelerating SCF by using derivative information to achieve superlinear convergence under suitable assumptions. For NEPv (1), applying Newton’s acceleration also requires care because the solution basis is not uniquely determined. Most existing Newton-type methods are derived for real-variable NEPv with X∈Rn,kX \in\mathbb{R}^{n,k} and H(X)∈Rn,nH(X) \in\mathbb{R}^{n,n}. A direct approach applies Newton’s method to the equations H(X)X−XΛ=0H(X)X-X\Lambda=0 and XTX=IkX^{T}X=I_k; see, e.g., [28, 73]. However, the unitary invariance (2) implies that a solution (X,Λ)(X,\Lambda) is not isolated, since (XQ,QTΛQ)(XQ,Q^{T}\Lambda Q) remains a solution for every unitary QQ, which leads to a singular Jacobian (derivative) operator that creates difficulties for standard Newton analysis and numerical solution [73]. To avoid this singularity, the approach in [40] replaces the orthogonality constraint XTX=IkX^{T}X=I_k by a normalization condition CTX=IkC^{T}X=I_k with a fixed matrix C∈Rn,kC \in\mathbb{R}^{n,k}. This formulation assumes that the NEPv satisfies H(XQ)=H(X)H(XQ)=H(X) for every nonsingular Q∈Rk,kQ \in\mathbb{R}^{k,k}, which is stronger than the unitary invariance (2). Along a different line, Riemannian Newton methods on the Grassmann manifold have been developed for NEPv arising from energy minimization, which removes the basis non-uniqueness by working with subspaces; see, e.g., [3, 78]. These derivations, however, rely on an underlying energy functional and differential-geometric tools. To our knowledge, a Newton acceleration for general NEPv (1), based on an elementary derivation and a correction operator that is nonsingular under suitable assumptions, has yet to be developed.

Contributions. The main contributions of this work are threefold:

  • For Anderson acceleration, we present three formulations that resolve the basis non-uniqueness of SCF iteration: (i) a parameterization-based framework that exploits the structure of H(X)H(X), (ii) an alignment-based AA, and (iii) an adaptation of Riemannian AA. The first extends the density-based approach to general NEPv, while the latter two work with n×kn \times k basis matrices without requiring a parameterization of H(X)H(X). The regularity properties of the associated SCF maps are established.

  • For Newton acceleration, we introduce the notion of a basis-aligned update to characterize local corrections to an approximate solution of the NEPv. This naturally removes basis non-uniqueness and leads to a Newton equation for correction through elementary linear algebra operations. It also provides a unified treatment to real and complex NEPv. Under suitable assumptions, we establish the nonsingularity of the correction operator and local quadratic convergence.

  • Through extensive numerical experiments on NEPv arising from applications in computational physics and machine learning, we confirm the theoretical findings and demonstrate the effectiveness and computational benefits of the proposed acceleration techniques.

Organization. The rest of the paper is organized as follows. Section 2 reviews Anderson acceleration and presents three variants for SCF, including parameterization-based, alignment-based, and Riemannian formulations of AA. Section 3 derives the basis-aligned Newton correction, establishes its local quadratic convergence, and discusses its implementation. Section 4 presents numerical experiments on problems from computational physics and data science. Concluding remarks and future directions are given in Section 5.

Notation. Throughout this paper, we use standard notation from matrix analysis. We denote by Rn,k\mathbb{R}^{n,k} and Cn,k\mathbb{C}^{n,k} the spaces of real and complex n×kn \times k matrices, respectively, with Rn=Rn,1\mathbb{R}^{n}=\mathbb{R}^{n,1} and Cn=Cn,1\mathbb{C}^{n}=\mathbb{C}^{n,1}. The set of complex orthonormal matrices is denoted by Un,k:={X∈Cn,k:XHX=Ik}\mathbb{U}_{n,k}:=\{X \in\mathbb{C}^{n,k}: X^{H}X=I_k\}. The Grassmann manifold of kk-dimensional subspaces of Cn\mathbb{C}^{n} (or Rn\mathbb{R}^{n} in the real case) is denoted by G(n,k)\mathcal{G}(n,k). For a matrix AA, ATA^{T} and AHA^{H} denote its transpose and conjugate transpose, respectively, range⁡(A)\operatorname{range}(A) denotes its column space, and vec⁡(A)\operatorname{vec}(A) denotes the vector obtained by stacking the columns of AA. For square AA, tr⁡(A)\operatorname{tr}(A) denotes its trace. The identity matrix of size nn is denoted by InI_n, with the subscript omitted when its size is clear. We use diag⁡(A)\operatorname{diag}(A) for the vector of diagonal entries of AA and Diag⁡(ν)\operatorname{Diag}(\nu) for the diagonal matrix with diagonal vector ν\nu. For a Hermitian A∈Cn,nA \in\mathbb{C}^{n,n}, its eigenvalues are ordered as λ1(A)≤⋯≤λn(A)\lambda_1(A) \leq\cdots\leq\lambda_n(A). For a block matrix A=[A1,…,Am]\mathcal{A}=[A_1,\ldots,A_m], and a vector β=(β1,…,βm)T∈Rm\beta=(\beta_1,\ldots,\beta_m)^T \in\mathbb{R}^{m}, we define the block product A⊙β:=∑j=1mβjAj\mathcal{A}\odot\beta:=\sum_{j=1}^{m}\beta_jA_j. Unless otherwise stated, ∥⋅∥\|\cdot\| denotes the spectral norm for matrices and ∥⋅∥F\|\cdot\|_{F} denotes the Frobenius norm. For complex matrices, we use the (real) Frobenius inner product ⟨A,B⟩:=Re⁡(tr⁡(AHB))\langle A,B\rangle:= \operatorname{Re}(\operatorname{tr}(A^{H}B)). Differentiability of functions with complex variables is understood with respect to the real and imaginary parts of their entries, and we write DF(X)[E]DF(X)[E] for the Fréchet derivative of FF at XX along the direction EE. We use the standard big-OO and little-oo notation: f=O(g)f = O(g) means ∣f∣≤C∣g∣|f| \le C|g| for some constant CC as g→0g \to0, and f=o(g)f = o(g) means ∣f∣/∣g∣→0|f|/|g| \to0. Other notations will be explained as needed.

Anderson Acceleration for NEPv

We consider Anderson Acceleration (AA) of the SCF iteration (3). Recall that applying AA to SCF requires care because of the unitary invariance (2) of the NEPv. Indeed, each iterate Xi+1X_{i+1} by the SCF (3) is not uniquely defined: For any unitary matrix Q∈Uk,kQ \in\mathbb{U}^{k,k}, the matrix Xi+1←Xi+1QX_{i+1} \leftarrow X_{i+1}Q is also a valid solution. Thus, at the level of basis matrices, SCF is a set-valued fixed-point iteration

Xi+1∈g(Xi),(4)X_{i+1} \in g(X_i), \tag*{(4)}

where

g(X):={Y:Y∈Un,k is an orthonormal eigenbasis for k smallest eigenvalues of H(X)}.(5)g(X) := \{Y : Y \in\mathbb{U}^{n,k}\text{ is an orthonormal eigenbasis for }k\text{ smallest eigenvalues of }H(X)\}. \tag*{(5)}

References

  1. [1]P.-A. Absil, R. Mahony, and R. Sepulchre, Riemannian Geometry of Grassmann Manifolds with a View on Algorithmic Computation, Acta Appl. Math., 80 (2004), pp. 199–220.DOI
  2. [2]——, Optimization Algorithms on Matrix Manifolds, Princeton University Press, 2008.DOI
  3. [3]R. Altmann, D. Peterseim, and T. Stykel, Riemannian Newton Methods for Energy Minimization Problems of Kohn–Sham Type, J. Sci. Comput., 101 (2024), pp. 1–25.DOI
  4. [4]M. R. Amini, N. Usunier, and C. Goutte, Learning from multiple partially observed views – an application to multiviewal text categorization, in Advances in Neural Information Processing Systems, Red Hook, NY, USA, 2009, Curran Associates Inc., pp. 28–36.
  5. [5]D. G. Anderson, Iterative Procedures for Nonlinear Integral Equations, J. ACM, 12 (1965), pp. 547–560.DOI
  6. [6]D. G. M. Anderson, Comments on “Anderson Acceleration, Mixing and Extrapolation”, Numer. Algorithms, 80 (2019), pp. 135–234.DOI
  7. [7]Z. Bai, R.-C. Li, and D. Lu, Sharp Estimation of Convergence Rate for Self-Consistent Field Iteration to Solve Eigenvector-Dependent Nonlinear Eigenvalue Problems, SIAM J. Matrix Anal. Appl., 43 (2022), pp. 301–327.DOI
  8. [8]Z. Bai and D. Lu, Variational Characterization of Monotone Nonlinear Eigenvector Problems and Geometry of Self-Consistent Field Iteration, SIAM J. Matrix Anal. Appl., 45 (2024), pp. 84–111.arxiv.org/abs/2211.05179
  9. [9]Z. Bai, D. Lu, and B. Vandereycken, Robust Rayleigh Quotient Minimization and Nonlinear Eigenvalue Problems, SIAM J. Sci. Comput., 40 (2018), pp. A3495–A3522.DOI
  10. [10]W. Bao and Q. Du, Computing the Ground State Solution of Bose–Einstein Condensates by a Normalized Gradient Flow, SIAM J. Sci. Comput., 25 (2004), pp. 1674–1697.arxiv.org/abs/cond-mat/0303241
  11. [11]W. Bao and W. Tang, Ground-state solution of Bose–Einstein condensate by directly minimizing the energy functional, J. Comput. Phys., 187 (2003), pp. 230–254.DOI
  12. [12]N. A. Barnafi and M. Lupo Pasini, Two-Level Sketching Alternating Anderson Acceleration for Complex Physics Applications, SIAM J. Sci. Comput., 48 (2026), pp. A2680–A2703.arxiv.org/abs/2505.08587
  13. [13]T. Bendokat, R. Zimmermann, and P.-A. Absil, A Grassmann manifold handbook: basic geometry and computational aspects, Adv. Comput. Math., 50 (2024), pp. 1–51.DOI
  14. [14]P. Benner, V. Mehrmann, A. Prajapati, and P. Sharma, Computation of structured stability radii for Dissipative-Hamiltonian systems, 2025. arXiv preprint arXiv:2511.14935.arxiv.org/abs/2511.14935
  15. [15]Å. Björck, Numerical methods for least squares problems, SIAM, 2 ed., 2024.
  16. [16]J.-L. Bretonnet, Basics of the density functional theory, AIMS Mater. Sci., 4 (2017), pp. 1372–1405.DOI
  17. [17]Y. Cai, L.-H. Zhang, Z. Bai, and R.-C. Li, On an Eigenvector-Dependent Nonlinear Eigenvalue Problem, SIAM J. Matrix Anal. Appl., 39 (2018), pp. 1360–1382.DOI
  18. [18]E. Cancès, G. Kemlin, and A. Levitt, Convergence Analysis of Direct Minimization and Self-Consistent Iterations, SIAM J. Matrix Anal. Appl., 42 (2021), pp. 243–274.
  19. [19]E. Cancès and C. Le Bris, On the convergence of SCF algorithms for the Hartree-Fock equations, ESAIM Math. Model. Numer. Anal., 34 (2000), pp. 749–774.
  20. [20]M. Chupin, M.-S. Dupuy, G. Legendre, and E. Séré, Convergence analysis of adaptive DIIS algorithms with application to electronic ground state calculations, ESAIM Math. Model. Numer. Anal., 55 (2021), pp. 2785–2825.arxiv.org/abs/2002.12850
  21. [21]P. Echenique and J. L. Alonso, A mathematical and computational review of Hartree–Fock SCF methods in quantum chemistry, Mol. Phys., 105 (2007), pp. 3057–3098.DOI
  22. [22]A. Edelman, T. A. Arias, and S. T. Smith, The Geometry of Algorithms with Orthogonality Constraints, SIAM J. Matrix Anal. Appl., 20 (1998), pp. 303–353.arxiv.org/abs/physics/9806030
  23. [23]T. Eirola, M. Huhtanen, and J. von Pfaler, Solution Methods for ℝ-Linear Problems in ℂⁿ, SIAM J. Matrix Anal. Appl., 25 (2003), pp. 804–828.
  24. [24]S. C. Eisenstat and H. F. Walker, Globally Convergent Inexact Newton Methods, SIAM J. Optim., 4 (1994), pp. 393–422.DOI
  25. [25]——, Choosing the Forcing Terms in an Inexact Newton Method, SIAM J. Sci. Comput., 17 (1996), pp. 16–32.
  26. [26]C. Evans, S. Pollock, L. G. Rebholz, and M. Xiao, A proof that Anderson acceleration improves the convergence rate in linearly converging fixed-point methods (but not in those converging quadratically), SIAM J. Numer. Anal., 58 (2020), pp. 788–810.arxiv.org/abs/1810.08455
  27. [27]H. Fang and Y. Saad, Two classes of multisecond methods for nonlinear acceleration, Numer. Linear Algebra Appl., 16 (2009), pp. 197–221.DOI
  28. [28]W. Gao, C. Yang, and J. C. Meza, Solving a Class of Nonlinear Eigenvalue Problems by Newton’s Method, Tech. Rep. LBNL-2187E, Ernest Orlando Lawrence Berkeley National Laboratory, Berkeley, CA (US), 2009.DOI
  29. [29]C. Garner, G. Lerman, and T. Zhang, Improved convergence factor of windowed Anderson acceleration for symmetric fixed-point iterations, SIAM J. Numer. Anal., 64 (2026), pp. 1891–1922.arxiv.org/abs/2311.02490
  30. [30]A. J. Garza and G. E. Scuseria, Comparison of self-consistent field convergence acceleration techniques, J. Chem. Phys., 137 (2012), p. 054110.DOI
  31. [31]C. L. Giles, K. D. Bollacker, and S. Lawrence, CiteSeer: an automatic citation indexing system, in Proceedings of the Third ACM Conference on Digital Libraries, New York, NY, USA, 1998, Association for Computing Machinery, pp. 89–98.DOI
  32. [32]G. H. Golub and C. F. Van Loan, Matrix Computations, Johns Hopkins University Press, Philadelphia, PA, 4 ed., 2013.
  33. [33]D. R. Hartree, The wave mechanics of an atom with a non-Coulomb central field. Part I. Theory and methods, Math. Proc. Cambridge Philos. Soc., 24 (1928), pp. 89–110.DOI
  34. [34]P. Henning and E. Jarlebring, The Gross–Pitaevskii Equation and Eigenvector Nonlinearities: Numerical Methods and Algorithms, SIAM Rev., 67 (2025), pp. 256–317.DOI
  35. [35]N. J. Higham, Functions of Matrices, Society for Industrial and Applied Mathematics, 2008.
  36. [36]X. Hu and W. Yang, Accelerating self-consistent field convergence with the augmented Roothaan–Hall energy function, J. Chem. Phys., 132 (2010), p. 054109.DOI
  37. [37]M. Huhtanen and A. Perämäki, Orthogonal polynomials of the ℝ-linear generalized minimal residual method, J. Approx. Theory, 167 (2013), pp. 220–239.arxiv.org/abs/1111.5167
  38. [38]J. Hull, A database for handwritten text recognition research, IEEE Trans. Pattern Anal. Mach. Intell., 16 (1994), pp. 550–554. available from https://www.csie.ntu.edu.tw/~cjlin/libs/ svmtools/ datasets/multiclass.html#usps.DOI
  39. [39]E. Jarlebring, S. Kvaal, and W. Michiels, An Inverse Iteration Method for Eigenvalue Problems with Eigenvector Nonlinearities, SIAM J. Sci. Comput., 36 (2014), pp. A1978–A2001.arxiv.org/abs/1212.0417
  40. [40]E. JARLEBRING AND P. UPADHYAYA, Implicit algorithms for eigenvector nonlinearities, Numer. Algorithms, 90 (2022), pp. 301–321.arxiv.org/abs/2002.12805
  41. [41]K. JBILOU, A. MESSAOUDI, AND H. SADOK, Global FOM and GMRES algorithms for matrix equations, Appl. Numer. Math., 31 (1999), pp. 49–63.DOI
  42. [42]I. T. JOLLIFFE, N. T. TRENDАFILOV, AND M. UDDIN, A Modified Principal Component Technique Based on the LASSO, J. Comput. Graph. Statist., 12 (2003), pp. 531–547.DOI
  43. [43]T. KATO, Perturbation Theory for Linear Operators, Classics in Mathematics, Springer-Verlag, Berlin, 2 ed., 1995.
  44. [44]C. T. KELLEY, Iterative methods for linear and nonlinear equations, SIAM, 1995.
  45. [45]G. P. KERKER, Efficient iteration scheme for self-consistent pseudopotential calculations, Phys. Rev. B, 23 (1981), pp. 3082–3084.DOI
  46. [46]K. N. KUDIN, G. E. SCUSERIA, AND E. CANCÈS, A black-box self-consistent field convergence algorithm: One step closer, J. Chem. Phys., 116 (2002), pp. 8255–8261.DOI
  47. [47]J. LI, K. CHENG, S. WANG, F. MORSTATTER, R. P. TREVINO, J. TANG, AND H. LIU, Feature Selection: A Data Perspective, ACM Comput. Surv., 50 (2018), pp. 1–45. Data available from https://jundongl.github.io/scikit-feature/.
  48. [48]R.-C. LI, A theory of the NEPv approach for optimization on the Stiefel manifold, Found. Comput. Math., 26 (2026), pp. 179–244.
  49. [49]R.-C. LI, L. WANG, L.-H. ZHANG, AND Z. BAI, NEPv approach for optimization on Stiefel manifold with the (2,1)-norm regularization, 2026. arXiv preprint arXiv:2609.26675.arxiv.org/abs/2609.26675
  50. [50]Z. LI AND C. BAO, Riemannian Anderson Mixing Methods for Minimizing C² Functions on Riemannian Manifolds, Math. Oper. Res., 51 (2026), pp. 905–937.arxiv.org/abs/2309.04091
  51. [51]X. LIU, X. WANG, Z. WEN, AND Y. YUAN, On the Convergence of the Self-Consistent Field Iteration in Kohn–Sham Density Functional Theory, SIAM J. Matrix Anal. Appl., 35 (2014), pp. 546–558.arxiv.org/abs/1302.6022
  52. [52]D. LU AND R.-C. LI, Locally unitarily invariantizable NEPv and convergence analysis of SCF, Math. Comp., 93 (2024), pp. 2291–2329.arxiv.org/abs/2212.14098
  53. [53]D. LU, A. PRAJAPATI, P. SHARMA, AND S. BORA, Eigenvalue Backward Errors of Rosenbrock Systems and Optimization of Sums of Rayleigh Quotients, SIAM J. Matrix Anal. Appl., 46 (2025), pp. 1301–1327.DOI
  54. [54]M. LUPO PASINI, Convergence analysis of Anderson-type acceleration of Richardson’s iteration, Numer. Linear Algebra Appl., 26 (2019), p. e2241.DOI
  55. [55]R. M. MARTIN, Electronic Structure: Basic Theory and Practical Methods, Cambridge University Press, 1 ed., 2004.
  56. [56]A. K. MCCALLUM, K. NIGAM, J. RENNIE, AND K. SEYMORE, Automating the construction of internet portals with machine learning, Inf. Retr., 3 (2000), pp. 127–163.
  57. [57]S. NENE, S. NAYAR, AND H. MURASE, Columbia Object Image Library (COIL-20), Tech. Rep. CUCS-005-96, Computer Vision Laboratory, Department of Computer Science, Columbia University, 1996. available from https://www.cs.columbia.edu/CAVE/software/softlib/coil-20.php.
  58. [58]P. PULAY, Convergence acceleration of iterative sequences. The case of SCF iteration, Chem. Phys. Lett., 73 (1980), pp. 393–398.DOI
  59. [59]P. PULAY, Improved SCF convergence acceleration, J. Comput. Chem., 3 (1982), pp. 556–560.DOI
  60. [60]C. C. J. ROOTHAAN, New developments in molecular orbital theory, Rev. Mod. Phys., 23 (1951), pp. 69–89.DOI
  61. [61]Y. Saad, A Flexible Inner-Outer Preconditioned GMRES Algorithm, SIAM J. Sci. Comput., 14 (1993), pp. 461–469.DOI
  62. [62]Y. Saad, Iterative Methods for Sparse Linear Systems, SIAM, Society for Industrial and Applied Mathematics, Philadelphia, PA, 2. ed. ed., 2007.
  63. [63]——, Acceleration Methods for Fixed-Point Iterations, Acta Numer., 34 (2025), pp. 805–890.arxiv.org/abs/2507.11746
  64. [64]V. R. Saunders and I. H. Hillier, A “Level-Shifting” method for converging closed shell Hartree–Fock wave functions, Int. J. Quantum Chem., 7 (1973), pp. 699–705.DOI
  65. [65]P. Sen, G. Namata, M. Bilgic, L. Getoor, B. Gallagher, and T. Eliassi-Rad, Collective Classification in Network Data, AI Mag., 29 (2008), pp. 93–106.DOI
  66. [66]L. Thøgersen, J. Olsen, D. Yeager, P. Jørgensen, P. Salek, and T. Helgaker, The trust-region self-consistent field method: Towards a black-box optimization in Hartree–Fock and Kohn–Sham theories, J. Chem. Phys., 121 (2004), pp. 16–27.pubmed.ncbi.nlm.nih.gov/15260518
  67. [67]A. Toth and C. T. Kelley, Convergence Analysis for Anderson Acceleration, SIAM J. Numer. Anal., 53 (2015), pp. 805–819.DOI
  68. [68]P. Upadhyaya, E. Jarlebring, and E. H. Rubensson, A Density Matrix Approach to the Convergence of the Self-Consistent Field Iteration, Numer. Algebra Control Optim., 11 (2021), pp. 99–115.arxiv.org/abs/1809.02183
  69. [69]H. F. Walker and P. Ni, Anderson Acceleration for Fixed-Point Iterations, SIAM J. Numer. Anal., 49 (2011), pp. 1715–1735.
  70. [70]L. Wang, J. Wang, and R.-C. Li, Sparse PCA via matrix (ℓ,1)-norm regularization with an application to feature selection, Results Appl. Math., 28 (2025), pp. 1–16.
  71. [71]L. Wang, L.-H. Zhang, and R.-C. Li, Trace ratio optimization with an application to multi-view learning, Math. Program., 201 (2023), pp. 97–131.arxiv.org/abs/2101.04292
  72. [72]Y. Wang and L. Chen, Multi-view fuzzy clustering with minimax optimization for effective clustering of data from multiple sources, Expert Syst. Appl., 72 (2017), pp. 457–466.arxiv.org/abs/1608.07005
  73. [73]T. Werner, An inexact Matrix-Newton method for solving NEPv, Linear Algebra Appl., 721 (2025), pp. 366–398. The 25th Conference of the International Linear Algebra Society.arxiv.org/abs/2311.09670
  74. [74]C. Yang, W. Gao, and J. C. Meza, On the Convergence of the Self-Consistent Field Iteration for a Class of Nonlinear Eigenvalue Problems, SIAM J. Matrix Anal. Appl., 30 (2009), pp. 1773–1788.DOI
  75. [75]L. Zhang and R. Li, Maximization of the sum of the trace ratio on the Stiefel manifold, I: Theory, Sci. China Math., 57 (2014), pp. 2495–2508.DOI
  76. [76]——, Maximization of the sum of the trace ratio on the Stiefel manifold, II: Computation, Sci. China Math., 58 (2015), pp. 1549–1566.DOI
  77. [77]Z. Zhang, Z. Zhai, and L. Li, Uniform Projection for Multi-View Learning, IEEE Trans. Pattern Anal. Mach. Intell., 39 (2017), pp. 1675–1689.DOI
  78. [78]Z. Zhao, Z.-J. Bai, and X.-Q. Jin, A Riemannian Newton Algorithm for Nonlinear Eigenvalue Problems, SIAM J. Matrix Anal. Appl., 36 (2015), pp. 752–774.DOI
  79. [79]P. Zhu and A. Knyazev, Angles between Subspaces and Their Tangents, J. Numer. Math., 21 (2013), pp. 325–340.DOI
  80. [80]H. Zou, T. Hastie, and R. Tibshirani, Sparse principal component analysis, J. Comput. Graph. Statist., 15 (2006), pp. 265–286.

Paper details

vlnc:W_2u27gfzd
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents