Pith. sign in

REVIEW 2 major objections 5 minor 60 references

Multi-block Bregman proximal alternating linearized minimization and its application to orthogonal nonnegative matrix factorization

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Two Bregman proximal alternating linearized algorithms are proven to converge to critical points for nonconvex nonsmooth multi-block problems, and applied with closed-form updates to penalized orthogonal nonnegative matrix factorization.

desk verdict Solid multi-block Bregman PALM theory, but the ONMF application overclaims because the penalized problem has spurious stationary points that are not orthogonal factorizations. read the letter →

arxiv 1908.01402 v2 pith:2KSQNXPL submitted 2019-08-04 math.OC cs.NAmath.NA

Masoud Ahookhosh , Le Thi Khanh Hien , Nicolas Gillis , Panagiotis Patrinos This is my paper · ORCID
classification math.OCcs.NAmath.NA MSC 90C2690C3065K05
keywords Bregmanproximalalternatinglinearizedminimizationmulti-blockrelativesmoothnessKurdyka-Łojasiewiczinequalityorthogonalnonnegativematrixfactorizationnonconvexnonsmoothoptimizationclosed-formupdatesconvergencerates
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces two block-coordinate algorithms, BPALM and its adaptive variant A-BPALM, for minimizing an objective made of a smooth nonconvex function $f$ plus block-separable nonsmooth nonconvex terms $g_i$. The central claim is that, under a multi-block relative smoothness condition that replaces Lipschitz-gradient knowledge with a Bregman distance, every bounded sequence generated by either algorithm has cluster points that are critical points; under the Kurdyka-Łojasiewicz inequality the whole sequence converges, with rates governed by the Łojasiewicz exponent. This matters because the framework is shown to cover orthogonal nonnegative matrix factorization: with a product-separable Bregman kernel the block subproblems admit closed-form solutions, giving what the authors describe as the first convergence-guaranteed algorithms for ONMF. The paper thus offers both a general convergence certificate for alternating Bregman methods and a concrete factorization problem where the certificate is actually used.

What carries the argument

The central object is the multi-block Bregman proximal alternating linearized mapping $$$T_i^{{h/\gamma_i}}$(x)=\arg\min_{z\in\mathbb{R}^{n_i}}\left\{\langle\nabla_i f(x),\,z-x_i\rangle+\frac{1}{\gamma_i}D_h(x+U_i(z-x_i),x)+g_i(z)\right\},$$ where $D_h$ is the Bregman distance of a multi-block kernel $h$ and $U_i$ selects the $i$-th block. The structural assumption doing the heavy lifting is multi-block relative smoothness: $f$ is $(L_1,\ldots,L_N)$-smooth relative to $h$ if $L_i h(x+U_i(z-x_i))-f(x+U_i(z-x_i))$ is convex in $z$ for every $x$ and $i$. This gives a blockwise majorization, the multi-block descent lemma, that makes the function decrease at every subproblem solve, and it replaces the usual Lipschitz moduli of partial gradients with kernel-dependent constants $L_i$. The same inequality, together with the optimality conditions of $T_i^{h/\gamma_i}$, yields the subgradient lower bound on iterate gaps that all subsequential and KL-based global convergence results rely on.

What would settle it

Exhibit a bounded sequence generated by Algorithm 1 that satisfies all assumptions of Theorem 3.4 but whose cluster set contains a point $x^\star$ with $0\notin\partial\phi(x^\star)$; for instance, engineer two-block nonconvex terms $g_i$ with flat regions to create a spurious accumulation point, and check the subgradient residual (3.6) along the orbit. Alternatively, enumerate exact solutions of the hard ONMF problem (4.2) for a small random matrix and check whether continuation applied to (4.6) returns a KKT point of the hard problem; failure would falsify the paper's implied penalty route.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is a convergence theorem for the Bregman proximal alternating linearized minimization scheme applied to problem (1.1). For a multi-block kernel $h$, the objective component $f$ is called $(L_1,\ldots,L_N)$-smooth relative to $h$ when each blockwise function $\varphi_i^x(z)=L_i h(x+U_i(z-x_i))-f(x+U_i(z-x_i))$ is convex; this yields a blockwise descent inequality without any Lipschitz constant for $\nabla_i f$. With step sizes $\gamma_i<1/L_i$, each block update decreases $\phi$ by a positive multiple of the block Bregman distance, and the optimality conditions of the subproblem provide a subgradient of $\phi$ at the new point whose norm is bounded by the iterate gap. Theorem 3.4 then gives subsequential convergence to critical points of bounded sequences, Theorem 3.8 gives global convergence and finite length under the Kurdyka-Łojasiewicz property, and Theorem 3.11 gives finite, linear, or sublinear rates depending on the Łojasiewicz exponent. For penalized ONMF, the kernel $h(U,V)=(\beta_1/2\|U\|_F^2+1)(\alpha_2/4\|V\|_F^4+\beta_2/2\|V\|_F^2+1)$ makes the objective multi-block relatively smooth, and the $U$- and $V$-subproblems are solved in closed form, the latter through the root of a scalar cubic.

Load-bearing premise

The load-bearing premise is that solving the penalized problem (4.6) with an increasing penalty parameter $\lambda$ is a faithful route to the original orthogonal NMF problem (4.2); the paper does not prove a stationary-point correspondence between the two problems, and its own Table 1 shows that larger initial $\lambda$ can worsen reconstruction error.

Editorial extensions

If this is right

  • Every bounded BPALM orbit has all its cluster points in the critical set of the composite objective, so stationarity can be certified without knowing partial-gradient Lipschitz constants.
  • Under the Kurdyka-Łojasiewicz property the whole sequence converges to a single critical point and the successive-iterate gaps are summable.
  • The Łojasiewicz exponent controls the speed: finite termination for $\theta=0$, linear convergence for $\theta\in(0,1/2]$, and sublinear convergence for $\theta\in(1/2,1)$.
  • For penalized ONMF, both block updates are closed-form operations, so each sweep of BPALM or A-BPALM is deterministic and cheap, which is what makes large-scale use plausible.
  • The adaptive backtracking variant A-BPALM inherits the subsequential and global convergence results while bounding the number of extra oracle calls per sweep.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper stops at the penalized ONMF objective; establishing a homotopy between stationary points of (4.6) and critical points of the exact problem (4.2) would turn the continuation procedure in Algorithm 3 from a heuristic into a certified method.
  • The multi-block relative smoothness setting is not tied to the ONMF kernel; it should transfer to other block factorization models, such as sparse NMF or nonnegative tensor factorization, once a sum- or product-separable kernel satisfying the blockwise convexity inequalities is identified.
  • The closed-form $V$-update uses Cardano's formula for a scalar cubic; a numerically guarded root solve of the same cubic would improve stability at scale without changing the convergence guarantees.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper introduces two multi-block proximal alternating linearized minimization algorithms, BPALM and A-BPALM, for minimizing f + sum_i g_i where f is (L_1,...,L_N)-smooth relative to a multi-block Bregman kernel h. The authors prove a sufficient decrease condition, summability of Bregman gaps, a subgradient lower bound, subsequential convergence to critical points, and global convergence with rates under a KL/Łojasiewicz-type assumption. They then instantiate the framework for the penalized orthogonal nonnegative matrix factorization (ONMF) formulation (4.6), derive closed-form block updates, and report numerical experiments on a synthetic data set and the Hubble telescope data set.

Significance. If the convergence results are correct, the multi-block relative smoothness framework is a useful generalization of PALM and B-PALM, and the closed-form updates for the penalized ONMF problem are practically valuable. The theoretical sections are largely a parameter-free derivation from stated assumptions: the relative smoothness constants in (4.5) are analytic bounds, and the KL-based arguments follow a standard template. The authors also make the code publicly available. However, the advertised application-level guarantee for ONMF is not supported by the theorems, which address only the penalized objective, and one key step in the subgradient lower bound needs repair. The core algorithmic idea is promising, but the manuscript currently overstates its scope.

major comments (2)
  1. [§4.1, Eqs. (4.2), (4.6), and Algorithm 3] The convergence guarantees proved in Section 3 apply to the penalized problem (4.6), not to the orthogonal NMF problem (4.2), and no theorem in the paper relates critical points of (4.6) to feasible points of (4.2). This is not a minor omission: the penalty has spurious stationary points for every λ>0, since the V-gradient of (λ/2)||I − VV^T||_F^2 is 2λ(VV^T V − V), which vanishes for any partial isometry V. For example, take nonnegative unit vectors u and w, set X = u w^T, U0 = [u, 0], V0 = [w; 0]; then (U0, V0) is a critical point of (4.6) for every λ>0 (both block gradients vanish), yet V0 V0^T = diag(1,0) ≠ I, so it is not feasible for (4.2). The same point is also a fixed point of the BPALM updates (4.7)–(4.8), since each subproblem is strictly convex and uniquely minimized there. Consequently the abstract's claim that BPALM and A-BPALM give rigorous convergence theory for ONMF is an overclaim; only the penalized formulation is covered. Algorithm 3's continuation over λ has no convergence analysis, and Table 1 shows that increasing the initial penalty can worsen Ferror, so the heuristic status of the route to (4.2) should be acknowledged and the claims reframed.
  2. [§3, Proposition 3.3, Eq. (3.6)] The vector G^{k+1}_i is defined using ∇_i f(x^{k,i}), but for i < N the point x^{k,i} differs from x^{k+1} in blocks i+1,...,N, so the claimed inclusion (G^{k+1}_1,...,G^{k+1}_N) ∈ ∂ϕ(x^{k+1}) is not established and is in general false. This is load-bearing because Theorem 3.4 and the subsequent KL-based results use this inclusion to pass to the limit and conclude 0 ∈ ∂ϕ(x*). The gap is repairable: replacing ∇_i f(x^{k,i}) by ∇_i f(x^{k+1}) yields a true subgradient, and the extra term ∇_i f(x^{k+1}) − ∇_i f(x^{k,i}) can be controlled by the same Lipschitz/block-displacement argument; but as written the proof is incomplete.
minor comments (5)
  1. [§4.1] The sentence 'The unconstrained version of the ONMF problem (4.2) is given by (4.6)' is misleading: (4.6) still contains nonnegativity constraints via indicator functions and a penalty term, so it is a penalized formulation rather than an unconstrained problem.
  2. [§4.2, Theorem 4.2, Eq. (4.9)] In the definition of τ2, the triple-bar norm '|||·|||' should be the Frobenius norm; please correct the notation.
  3. [§4.2, Table 1] The caption says 'the smallest number of Ferror and Ferror are displayed in bold' but should refer to Ferror and Oerror; moreover, the bold entries do not always appear to be the smallest in their columns (for example, the BPALM row with λ=1 has Ferror 6.75e-2, which is not the minimum).
  4. [§4.2] The text refers to 'A-BPALM1 and A-BPALM1' when describing the two adaptive variants; the second should be A-BPALM2.
  5. [§3, Proposition 3.3 proof] In the displayed inequality for i=1,...,N−1, the summation index i is reused both as a fixed block index and as a summation index; using separate indices would improve clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the convergence theory is derived from stated assumptions, and the ONMF constants are analytic bounds, not fitted values.

full rationale

The claimed convergence results (Theorems 3.4, 3.8, and 3.11) are derived in-chain from Assumption I, multi-block relative smoothness (Definition 2.4), the Bregman proximal alternating inequality (Lemma 2.9), and the subgradient lower bound (Proposition 3.3), with standard external KL facts (Fact 3.6 from [21], Fact 3.7 from [23]) used as premises. No quantity that is later called a prediction is fitted to data: the ONMF relative smoothness constants in (4.5) are analytic bounds obtained from the Hessian comparison in Proposition 4.1, and the closed-form updates (4.7)-(4.8) are derived by solving the Bregman proximal subproblems. The only self-citation, [2], supplies standard Bregman kernel and prox-boundedness vocabulary and is not load-bearing. The abstract's claim that BPALM and A-BPALM are the first algorithms with rigorous convergence theory for ONMF is stronger than what is proven: the theorems apply to the penalized formulation (4.6), and Algorithm 3's outer continuation is heuristic, with no theorem relating stationary points of (4.6) to feasible points of the original ONMF problem (4.2). That is an overclaim or correctness gap, not a circularity, because the convergence argument for the penalized objective does not presuppose the desired ONMF conclusion.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

No new physical or formal entities are postulated. Multi-block kernel functions and Bregman proximal alternating mappings are new definitions within the framework, but they are mathematical constructs with no independent falsifiable handles.

free parameters (4)
  • relative smoothness constants L1, L2 = For ONMF with α2=β1=β2=1: L1=2, L2=12; step sizes γ_i=1/L_i minus machine epsilon
    The descent lemma and convergence proofs require γ_i < 1/L_i. L_i are analytic upper bounds from Proposition 4.1, not fitted to data, but they determine step sizes and the user must know or adapt them.
  • kernel parameters α2, β1, β2 = set to 1 in Section 4.2
    They define the Bregman kernel h in (4.4). The choice is by hand; different values change step-size bounds and the resulting method.
  • backtracking hyperparameters ν1 and L_i^0 = ν1=2, L_i^0=0.01L_i for A-BPALM1 and 0.1L_i for A-BPALM2
    User-chosen values in Algorithm 2; they affect oracle-call counts and practical behavior, not the convergence conclusion.
  • penalty parameter λ and continuation factor c = synthetic: initial λ=10, c=1.5; Hubble: initial λ in {1,10,100,1000,10000}, c=1.5
    These control the penalized ONMF objective and the continuation schedule. They are experimental choices rather than fitted constants.
assumptions (6)
  • domain assumption Assumption I: f is (L1,...,LN)-smooth relative to h, each g_i is proper lsc, h is a multi-block Legendre kernel, argmin φ is nonempty, and dom φ ⊆ int dom h.
    This is the standing structural assumption in Section 1 and throughout; if it fails, the descent lemma and the Bregman proximal mapping properties do not hold.
  • domain assumption The generated sequence is bounded (Theorems 3.4 and 3.8).
    Boundedness is assumed rather than derived; Remark 3.9 notes it follows when level sets are bounded, but the theorem statement includes it as a hypothesis.
  • domain assumption h is locally multi-block strongly convex (Theorem 3.4) or globally multi-block strongly convex (Theorem 3.8) so that vanishing Bregman distances imply vanishing iterate gaps.
    Strong convexity is used in the proof to translate Dh(u,v)→0 into ||x^{k+1}-x^k||→0.
  • domain assumption φ satisfies the Kurdyka-Łojasiewicz property (Theorem 3.8) or the Łojasiewicz-type inequality (3.15) (Theorem 3.11).
    These are standard hypotheses for global convergence and rates in nonconvex optimization; they are not proved for ONMF.
  • standard math Uniformized KL property (Fact 3.6, [21, Lemma 6]) and summability lemma (Fact 3.7, [23, Lemma 2.3]) are valid.
    These external lemmas are imported without proof and used in the global convergence and rate arguments.
  • domain assumption For bounded sets, ∇_i f and ∇_i h have Lipschitz moduli \hat L_i and ~ L_i, used in Proposition 3.3.
    The subgradient lower bound relies on this Lipschitz condition; the paper assumes these moduli exist on the bounded neighborhoods where iterates lie.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-block Bregman proximal alternating linearized minimization and its application to orthogonal nonnegative matrix factorization." pith.science (2026). https://pith.science/paper/2KSQNXPL

@misc{pith2026190801402,
  author       = {Pith},
  title        = {Pith review of: Multi-block Bregman proximal alternating linearized minimization and its application to orthogonal nonnegative matrix factorization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2KSQNXPL}},
  note         = {Machine review of arXiv:1908.01402}
}
read the original abstract

We introduce and analyze BPALM and A-BPALM, two multi-block proximal alternating linearized minimization algorithms using Bregman distances for solving structured nonconvex problems. The objective function is the sum of a multi-block relatively smooth function (i.e., relatively smooth by fixing all the blocks except one) and block separable (nonsmooth) nonconvex functions. It turns out that the sequences generated by our algorithms are subsequentially convergent to critical points of the objective function, while they are globally convergent under KL inequality assumption. Further, the rate of convergence is further analyzed for functions satisfying the {\L}ojasiewicz's gradient inequality. We apply this framework to orthogonal nonnegative matrix factorization (ONMF) that satisfies all of our assumptions and the related subproblems are solved in closed forms, where some preliminary numerical results is reported.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 59 canonical work pages

  1. [1]

    M. A hookhosh, Accelerated first-order methods for large-scale convex optimization: nearly optimal com- plexity under strong convexity, Mathematical Methods of Operations Research, 89 (2019), pp. 319–353

    work page 2019

  2. [2]

    A Bregman forward-backward linesearch algorithm for nonconvex composite optimization: superlinear convergence to nonisolated local minima

    M. A hookhosh, A. Themelis, and P. Patrinos, Bregman forward-backward splitting for nonconvex composite optimization: superlinear convergence to nonisolated critical points, arXiv:1905.11904, (2019)

    work page Pith review arXiv 2019

  3. [3]

    F. J. A. A rtacho, R. M. F leming, and P. T. Vuong, Accelerating the DC algorithm for smooth functions , Mathematical Programming, 169 (2018), pp. 95–118

    work page 2018

  4. [4]

    A ttouch, J

    H. A ttouch, J. B olte, P. Redont, and A. Soubeyran, Alternating proximal algorithms for weakly coupled convex minimization problems. applications to dynamical games and PDE’s, Journal of Convex Analysis, 15 (2008), p. 485

    work page 2008

  5. [5]

    A ttouch, J

    H. A ttouch, J. B olte, P. R edont, and A. S oubeyran, Proximal alternating minimization and projection methods for nonconvex problems: An approach based on the Kurdyka-Łojasiewicz inequality, Mathematics of Operations Research, 35 (2010), pp. 438–457

    work page 2010

  6. [6]

    A ttouch, J

    H. A ttouch, J. B olte, and B. F. S v aiter, Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward-backward splitting, and regularized Gauss-Seidel methods, Math- ematical Programming, 137 (2013), pp. 91–129

    work page 2013

  7. [7]

    A ttouch, P

    H. A ttouch, P. Redont, and A. Soubeyran, A new class of alternating proximal minimization algorithms with costs-to-move, SIAM Journal on Optimization, 18 (2007), pp. 1061–1081

    work page 2007

  8. [8]

    A ttouch and A

    H. A ttouch and A. Soubeyran, Inertia and reactivity in decision making as cognitive variational inequali- ties, Journal of Convex Analysis, 13 (2006), p. 207

    work page 2006

Show all 60 references
  1. [9]

    A uslender, Optimisation méthodes numériques

    A. A uslender, Optimisation méthodes numériques. 1976, Mason, Paris, (1976)

    1976

  2. [10]

    H. H. B auschke, J. Bolte, J. Chen, M. Teboulle, and X. Wang, On linear convergence of non-Euclidean gra- dient methods without strong convexity and Lipschitz gradient continuity , Journal of Optimization Theory and Applications, (2019), pp. 1–20

    2019

  3. [11]

    H. H. B auschke, J. B olte, and M. Teboulle, A descent lemma beyond Lipschitz gradient continuity: first- order methods revisited and applications, Mathematics of Operations Research, 42 (2016), pp. 330–348

    2016

  4. [12]

    H. H. B auschke, M. N. D ao, and S. B. L indstrom, Regularizing with Bregman–Moreau envelopes, SIAM Journal on Optimization, 28 (2018), pp. 3208–3228

    2018

  5. [13]

    B eck, E

    A. B eck, E. Pauwels, and S. Sabach, The cyclic block conditional gradient method for convex optimization problems, SIAM Journal on Optimization, 25 (2015), pp. 2024–2049

    2015

  6. [14]

    B eck, S

    A. B eck, S. S abach, and M. T eboulle, An alternating semiproximal method for nonconvex regularized structured total least squares problems , SIAM Journal on Matrix Analysis and Applications, 37 (2016), pp. 1129–1150

    2016

  7. [15]

    B eck and M

    A. B eck and M. Teboulle, A fast iterative shrinkage-thresholding algorithm for linear inverse problems , SIAM Journal on Imaging Sciences, 2 (2009), pp. 183–202

    2009

  8. [16]

    B eck and L

    A. B eck and L. Tetruashvili, On the convergence of block coordinate descent type methods, SIAM journal on Optimization, 23 (2013), pp. 2037–2060

    2013

  9. [17]

    D. P. B ertsekas and J. N. Tsitsiklis, Parallel and Distributed Computation: Numerical Methods, Prentice- Hall, Inc., 1989. Multi-block Bregman proximal alternating linearized minimization 23

    1989

  10. [18]

    B olte, A

    J. B olte, A. Daniilidis, and A. Lewis, The Łojasiewicz inequality for nonsmooth subanalytic functions with applications to subgradient dynamical systems, SIAM Journal on Optimization, 17 (2007), pp. 1205–1223

    2007

  11. [19]

    B olte, A

    J. B olte, A. Daniilidis, A. Lewis, and M. Shiota, Clarke subgradients of stratifiable functions, SIAM Journal on Optimization, 18 (2007), pp. 556–572

    2007

  12. [20]

    B olte, A

    J. B olte, A. Daniilidis, O. L ey, and L. Mazet, Characterizations of Łojasiewicz inequalities: subgradient flows, talweg, convexity, Transactions of the American Mathematical Society, 362 (2010), pp. 3319–3363

    2010

  13. [21]

    B olte, S

    J. B olte, S. S abach, and M. T eboulle, Proximal alternating linearized minimization for nonconvex and nonsmooth problems, Mathematical Programming, 146 (2014), pp. 459–494

    2014

  14. [22]

    B olte, S

    J. B olte, S. S abach, M. T eboulle, and Y . Vaisbourd, First order methods beyond convexity and Lipschitz gradient continuity with applications to quadratic inverse problems , SIAM Journal on Optimization, 28 (2018), pp. 2131–2151

    2018

  15. [23]

    R. I. B o¸t and E. R. C setnek, An inertial Tseng’s type proximal algorithm for nonsmooth and nonconvex optimization problems, Journal of Optimization Theory and Applications, 171 (2016), pp. 600–616

    2016

  16. [24]

    R. I. B ot, E. R. C setnek, and D.-K. Nguyen, A proximal minimization algorithm for structured nonconvex and nonsmooth problems, SIAM Journal on Optimization, 29 (2019), pp. 1300–1328

    2019

  17. [25]

    B outsidis and E

    C. B outsidis and E. Gallopoulos, Svd-based initialization: A head start for nonnegative matrix factoriza- tion, Pattern Recognition, 41 (2008), pp. 1350–1362

    2008

  18. [26]

    L. M. B regman, The relaxation method of finding the common point of convex sets and its application to the solution of problems in convex programming , USSR Computational Mathematics and Mathematical Physics, 7 (1967), pp. 200–217

    1967

  19. [27]

    C hen and M

    G. C hen and M. Teboulle, Convergence analysis of a proximal-like minimization algorithm using Bregman functions, SIAM Journal on Optimization, 3 (1993), pp. 538–543

    1993

  20. [28]

    C ichocki, R

    A. C ichocki, R. Z dunek, A. H. P han, and S.-i. A mari, Nonnegative Matrix and Tensor Factorizations: Applications to Exploratory Multi-way Data Analysis and Blind Source Separation , John Wiley & Sons, 2009

    2009

  21. [29]

    P. L. C ombettes and J.-C. Pesquet, Stochastic quasi-Fejér block-coordinate fixed point iterations with ran- dom sweeping, SIAM Journal on Optimization, 25 (2015), pp. 1221–1248

    2015

  22. [30]

    F ercoq and P

    O. F ercoq and P. Bianchi, A coordinate-descent primal-dual algorithm with large step size and possibly nonseparable functions, SIAM Journal on Optimization, 29 (2019), pp. 100–134

    2019

  23. [31]

    X. F u, K. H uang, N. D. S idiropoulos, and W.-K. Ma, Nonnegative matrix factorization for signal and data analytics: Identifiability, algorithms, and applications, IEEE Signal Processing Magazine, 36 (2019), pp. 59–80

    2019

  24. [32]

    G illis, The why and how of nonnegative matrix factorization, Regularization, Optimization, Kernels, and Support Vector Machines, 12 (2014), pp

    N. G illis, The why and how of nonnegative matrix factorization, Regularization, Optimization, Kernels, and Support Vector Machines, 12 (2014), pp. 257–291

    2014

  25. [33]

    G rippo and M

    L. G rippo and M. Sciandrone, On the convergence of the block nonlinear Gauss–Seidel method under convex constraints, Operations Research Letters, 26 (2000), pp. 127–136

    2000

  26. [34]

    H anzely and P

    F. H anzely and P. R icht´arik, Fastest rates for stochastic mirror descent methods , arXiv preprint arXiv:1803.07374, (2018)

    2018 arXiv

  27. [35]

    H anzely, P

    F. H anzely, P. Richtarik, and L. Xiao, Accelerated Bregman proximal gradient methods for relatively smooth convex optimization, arXiv preprint arXiv:1808.03045, (2018)

    2018 arXiv

  28. [36]

    K an and W

    C. K an and W. Song, The Moreau envelope function and proximal mapping in the sense of the Bregman distance, Nonlinear Analysis: Theory, Methods & Applications, 75 (2012), pp. 1385 – 1399

    2012

  29. [37]

    K urdyka, On gradients of functions definable in o-minimal structures , Annales de l’institut Fourier, 48 (1998), pp

    K. K urdyka, On gradients of functions definable in o-minimal structures , Annales de l’institut Fourier, 48 (1998), pp. 769–783

    1998

  30. [38]

    L atafat, N

    P. L atafat, N. M. F reris, and P. Patrinos, A new randomized block-coordinate primal-dual proximal algo- rithm for distributed optimization, IEEE Transactions on Automatic Control, (2019)

    2019

  31. [39]

    L atafat, A

    P. L atafat, A. Themelis, and P. Patrinos, Block-coordinate and incremental aggregated nonconvex proximal gradient methods: a unified view, arXiv preprint arXiv:1906.10053, (2019)

    2019 arXiv

  32. [40]

    Q. L i, Z. Zhu, G. Tang, and M. B. Wakin, Provable Bregman-divergence based methods for nonconvex and non-Lipschitz problems, arXiv preprint arXiv:1904.09712, (2019)

    2019 arXiv

  33. [41]

    Ł ojasiewicz, Une propriété topologique des sous-ensembles analytiques réels, Les équations aux dérivées partielles, (1963), pp

    S. Ł ojasiewicz, Une propriété topologique des sous-ensembles analytiques réels, Les équations aux dérivées partielles, (1963), pp. 87–89

    1963

  34. [42]

    Ł ojasiewicz, Sur la géométrie semi- et sous- analytique , Annales de l’institut Fourier, 43 (1993), pp

    S. Ł ojasiewicz, Sur la géométrie semi- et sous- analytique , Annales de l’institut Fourier, 43 (1993), pp. 1575–1595

    1993

  35. [43]

    H. L u, R. M. Freund, and Y . Nesterov, Relatively smooth convex optimization by first-order methods, and applications, SIAM Journal on Optimization, 28 (2018), pp. 333–354

    2018

  36. [44]

    M. C. M ukkamala, P. Ochs, T. Pock, and S. Sabach, Convex-concave backtracking for inertial Bregman proximal gradient algorithms in non-convex optimization, arXiv preprint arXiv:1904.03537, (2019)

    2019 arXiv

  37. [45]

    Nesterov, Efficiency of coordinate descent methods on huge-scale optimization problems, SIAM Journal on Optimization, 22 (2012), pp

    Y . Nesterov, Efficiency of coordinate descent methods on huge-scale optimization problems, SIAM Journal on Optimization, 22 (2012), pp. 341–362

    2012

  38. [46]

    Nesterov, Gradient methods for minimizing composite functions , Mathematical Programming, 140 (2013), pp

    Y . Nesterov, Gradient methods for minimizing composite functions , Mathematical Programming, 140 (2013), pp. 125–161. 24 M. AHOOKHOSH, L.T.K. HIEN, N. GILLIS, AND P. PATRINOS

    2013

  39. [47]

    V . P. Pauca, J. Piper, and R. J. Plemmons, Nonnegative matrix factorization for spectral data analysis, Linear Algebra and its Applications, 416 (2006), pp. 29 – 47

    2006

  40. [48]

    P ock and S

    T. P ock and S. Sabach, Inertial proximal alternating linearized minimization (iPALM) for nonconvex and nonsmooth problems, SIAM Journal on Imaging Sciences, 9 (2016), pp. 1756–1787

    2016

  41. [49]

    P ompili, N

    F. P ompili, N. Gillis, P.-A. Absil, and F. Glineur, Two algorithms for orthogonal nonnegative matrix fac- torization with application to clustering, Neurocomputing, 141 (2014), pp. 15–25

    2014

  42. [50]

    R aza viy ayn, M

    M. R aza viy ayn, M. Hong, and Z.-Q. Luo, A unified convergence analysis of block successive minimization methods for nonsmooth optimization, SIAM Journal on Optimization, 23 (2013), pp. 1126–1153

    2013

  43. [51]

    R icht´arik and M

    P. R icht´arik and M. Tak´aˇc, Iteration complexity of randomized block-coordinate descent methods for min- imizing a composite function, Mathematical Programming, 144 (2014), pp. 1–38

    2014

  44. [52]

    R. T. R ockafellar and R. J.-B. Wets, Variational Analysis, vol. 317, Springer Science & Business Media, 2011

    2011

  45. [53]

    S hefi and M

    R. S hefi and M. Teboulle, On the rate of convergence of the proximal alternating linearized minimization algorithm for convex problems, EURO Journal on Computational Optimization, 4 (2016), pp. 27–46

    2016

  46. [54]

    M. K. T am, Regularity properties of non-negative sparsity sets, Journal of Mathematical Analysis and Ap- plications, 447 (2017), pp. 758–777

    2017

  47. [55]

    T eboulle, A simplified view of first order methods for optimization, Mathematical Programming, (2018), pp

    M. T eboulle, A simplified view of first order methods for optimization, Mathematical Programming, (2018), pp. 1–30

    2018

  48. [56]

    T hemelis, L

    A. T hemelis, L. S tella, and P. Patrinos, Forward-backward envelope for the sum of two nonconvex func- tions: Further properties and nonmonotone linesearch algorithms , SIAM Journal on Optimization, 28 (2018), pp. 2274–2303

    2018

  49. [57]

    T seng, Convergence of a block coordinate descent method for nondifferentiable minimization, Journal of OptimizationTheory and Applications, 109 (2001), pp

    P. T seng, Convergence of a block coordinate descent method for nondifferentiable minimization, Journal of OptimizationTheory and Applications, 109 (2001), pp. 475–494

    2001

  50. [58]

    T seng and S

    P. T seng and S. Yun, A coordinate gradient descent method for nonsmooth separable minimization, Mathe- matical Programming, 117 (2009), pp. 387–423

    2009

  51. [59]

    V an den Dries, Tame Topology and o-Minimal Structures, vol

    L. V an den Dries, Tame Topology and o-Minimal Structures, vol. 248, Cambridge university press, 1998

    1998

  52. [60]

    W ang, X

    X. W ang, X. Y uan, S. Z eng, J. Z hang, and J. Z hou, Block coordinate proximal gradi- ent method for nonconvex optimization problems: convergence analysis , http: //www.optimization- online.org/DB_HTML/2018/04/6573.html, (2018)

    2018

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.