Pith. sign in

REVIEW 4 major objections 5 minor 47 references

Sample Complexity of Branch-length Estimation by Maximum Likelihood

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read The paper proves that, deep in the Kesten–Stigum regime with polynomially many samples on a balanced tree, the CFN empirical log-likelihood is strongly concave on a box around the true branch lengths, so cyclic coordinate maximization…

desk verdict New finite-sample landscape and coordinate-descent convergence results for CFN branch-length MLE, but Theorem 3.2's universal sample-complexity bound does not follow from the proof; the gap is real and should be fixed. read the letter →

arxiv 2507.22038 v1 pith:7P4TO7VB submitted 2025-07-29 stat.CO

classification stat.CO MSC 62F1262M05
keywords branch-lengthestimationmaximumlikelihoodnon-concavelandscapeCavender-Farris-NeymanmodelcoordinatemaximizationKesten-Stigumreconstructionregimestrongconcavitysamplecomplexity
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 asks why simple coordinate maximization works so well in practice for estimating branch lengths in phylogenetics, even though the likelihood landscape is non-concave and can have many critical points. It argues that, deep in the Kesten-Stigum reconstruction regime (small mutation probabilities), the empirical log-likelihood of the two-state symmetric Cavender-Farris-Neyman (CFN) model is, with high probability, strongly concave and smooth on a fixed-size box around the true parameters, provided the number of samples $m$ is at least $(C/\delta)^{\mathrm{diam}(T)+8}\log(1/\varepsilon)$. On that event the confined maximum likelihood estimator is unique, lies within $O(\sqrt{|E|/m}\log(|E|/\varepsilon))$ of the truth, and the standard cyclic coordinate-maximization algorithm converges to it exponentially fast with a rate independent of the tree and of $\delta$, assuming the initializer is $O(\delta)$ from the truth. The contribution is a first rigorous explanation of the empirical success of hill-climbing on phylogenetic likelihoods, within a semi-global good region.

What carries the argument

The load-bearing object is the magnetization $Z_u$ of a node $u$: the posterior bias toward spin $+1$ given the observed spins at the leaves of a descendant subtree. Magnetizations satisfy the recursion $q(s,t) = (s+t)/(1+st)$, and the gradient of the log-likelihood along edge $e=\{x,y\}$ is $Z_x Z_y / (1 + Z_x Z_y \hat\theta_e)$. These formulas yield explicit Hessian entries whose denominators scale as $\Theta(\delta^2)$ and third-order derivatives bounded by $O(\mathrm{diam}(T)\,\delta^{-O(\mathrm{diam}(T))})$. Those bounds supply the Lipschitz regularity and uniform spectral-norm control needed to apply a uniform matrix Bernstein inequality, transferring the population strong-concavity theorem to the empirical Hessian uniformly over the box. A multi-block coordinate minimization lemma then converts strong concavity plus smoothness into the exponential convergence of cyclic coordinate maximization.

What would settle it

Fix a balanced binary tree with a small number of leaves, choose a small $\delta$ and a true parameter $\theta^*$ with mutation probabilities in $[c_3\delta, C_3\delta]$, draw $m=(C_5/\delta)^{\mathrm{diam}(T)+8}\log(1/\varepsilon)$ samples, and compute the empirical Hessian eigenvalues on a fine grid over $\widehat\Theta_0(\delta)$; any eigenvalue above $-C_6\delta^{-1}$ or below $-\tilde C_6\delta^{-1}$ would falsify Theorem 3.2. Similarly, running Algorithm 1 from initializers at L2 distance exactly $O(\delta)$ but pointed outside the box and observing escape from $\widehat\Theta_0(\delta)$ with non-negligible probability would falsify the confinement lemma underlying Theorem 3.4.

Watch

Extended reading notes

Core claim

In the paper's own terms, let $\theta^*$ be the true edge parameters in the box $\Theta_0(\delta)$, with mutation probabilities between $c_3\delta$ and $C_3\delta$, and let $\hat H$ be the Hessian of the $m$-sample log-likelihood. The central claim is that for all sufficiently small $\delta$ and $m \ge (C_5/\delta)^{\mathrm{diam}(T)+8}\log(1/\varepsilon)$, with probability at least $1-\varepsilon$, all eigenvalues of $\hat H(\theta)$ over the larger box $\widehat\Theta_0(\delta)$ lie between $-\tilde C_6 \delta^{-1}$ and $-C_6 \delta^{-1}$ (Theorem 3.2). Thus the empirical likelihood landscape is $\Theta(\delta^{-1})$-strongly concave and smooth there. Consequently the empirical maximizer $\hat\theta^*$ is unique with high probability and satisfies $\|\theta^*-\hat\theta^*\| \le C_7\sqrt{|E|/m}\,\log(|E|/\varepsilon)$ (Theorem 3.3), and cyclic coordinate maximization initialized at distance $O(\delta)$ from $\theta^*$ converges geometrically with universal contraction factor $1 - C_6/\tilde C_6$ (Theorem 3.4). The total error separates into a statistical term of order $O(1/\sqrt{m})$ and a computational term that shrinks by a universal factor each coordinate sweep.

Load-bearing premise

The results require a starting point within $O(\delta)$ of the true parameters, and the paper does not construct or justify such an initializer; if that input is not available, the exponential-convergence guarantee may never engage.

Editorial extensions

If this is right

  • For balanced trees, the sample complexity bound is polynomial in the number of leaves, so the strong-concavity guarantee survives as the tree grows.
  • The statistical error $O(\sqrt{|E|/m}\log(|E|/\varepsilon))$ and the computational error are cleanly separated, and a desired computational tolerance needs only $O(\log(1/\varepsilon))$ coordinate sweeps.
  • Uniqueness of the maximizer inside the box shows that known non-uniqueness examples for the global CFN likelihood do not obstruct estimation near the true parameter in this regime.
  • The theorem provides the first finite-sample regularity guarantee for this likelihood landscape, giving theoretical support to the practical success of coordinate-ascent phylogenetic likelihood methods.

Reading between the lines

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

  • If a $\delta$-close initializer can be constructed from cheap distance-based estimators or other preprocessing, the result would upgrade to an end-to-end polynomial-time algorithm; the paper leaves that initializer construction as the main open step.
  • The three-step template used here, population strong concavity plus Lipschitz Hessian entries plus uniform matrix Bernstein, should transfer to other latent-tree and mixture-model maximum likelihood problems with benign non-concavity; only the magnetization algebra would need to change.
  • The guarantee is semiglobal, not global: it explains why local optimization succeeds once inside the good box, but leaves open how an arbitrary initialization finds that box.
  • Because smaller $\delta$ strengthens the curvature scale $\delta^{-1}$ while simultaneously increasing the sample requirement $(C/\delta)^{\mathrm{diam}(T)+8}$, the theorem quantifies an explicit trade-off between signal strength and the amount of data needed to certify the landscape.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper analyzes maximum-likelihood branch-length estimation under the Cavender-Farris-Neyman model on a known binary tree. It claims that, deep inside the Kesten-Stigum reconstruction regime, with polynomially many samples (polynomial in the number of leaves when the tree is balanced), the empirical log-likelihood is strongly concave and smooth on an L-infinity box around the true parameter with high probability (Theorem 3.2). It further claims that the MLE over this box is O(1/sqrt(m))-consistent (Theorem 3.3) and that cyclic coordinate maximization, initialized within O(delta) of the truth, converges exponentially fast to the MLE (Theorem 3.4). The proofs combine population-landscape results from a companion paper (Clancy et al. 2025a), a uniform matrix Bernstein inequality, Berry-Esseen bounds, and a block-coordinate descent convergence lemma.

Significance. If the stated results could be established as written, they would be a valuable first rigorous explanation of the practical success of coordinate-ascent heuristics such as PHYML for phylogenetic likelihood maximization, and they would provide a transferable template for analyzing non-concave empirical likelihood landscapes. The paper is transparent about the restrictive initialization assumption in Theorem 3.4 and about its reliance on prior work for the population landscape. The analytic formulas for gradient and Hessian via magnetizations are explicit and potentially reusable. However, as detailed below, the current proof does not support the sample-complexity statements as formulated, and several concentration steps in the appendix need correction.

major comments (4)
  1. [Appendix C, Theorem C.1, eqs. (31)-(34)] The claimed sample complexity m >= (C31/delta)^{diam(T)+8} log(1/epsilon) does not follow from the displayed sufficient bound. Equation (34) has a prefactor |E|^2 (eC16/delta)^{diam(T)+8} and a logarithmic factor |E| log( 2|E|^{1/2}(C4-c4)delta (1 + 16 diam(T)(2c4 delta)^{-4diam(T)-2}) ), which behaves like |E|^3 diam(T) log(1/delta) as delta -> 0. For a fixed binary tree, the ratio of the proven m to the claimed m therefore diverges as delta -> 0, so no delta-independent and tree-independent constant C31 can absorb the delta-dependent logarithmic factor. The statements of Theorems 3.3 and 3.4 inherit this gap because they invoke Theorem 3.2.
  2. [Appendix C, display (33)] The application of Lemma 4.5 omits the factor m|E|^2 from the covering-number prefactor. With the Lipschitz constant L <= 4 diam(T)(2c4 delta)^{-4diam(T)-2} |E|^2 and t=1, n=m, Lemma 4.5 gives (1 + 4 n L / t)^p = (1 + 16 m |E|^2 diam(T)(2c4 delta)^{-4diam(T)-2})^{|E|}, not the displayed (1 + 16 diam(T)(2c4 delta)^{-4diam(T)-2})^{|E|}. This missing factor affects the derivation of a closed-form sufficient sample size and should be corrected before the sample-complexity claim can be assessed.
  3. [Theorem 3.3 and Theorem C.2, eq. (7), eq. (37)] The stated sample-size condition is inconsistent. The main text claims m = Omega(|E|^2/epsilon), while Theorem C.2 requires m >= |E|^2/(4 C19^6 c4^6 delta^3 epsilon). Moreover, even the appendix condition is insufficient: the Berry-Esseen term 3|E|/(8 C19^3 c4^3 delta^{3/2} sqrt(m)) in (37) is made at most epsilon only when m is of order |E|^2/(delta^3 epsilon^2); the displayed m gives only a bound of order sqrt(epsilon). Thus the stated consistency sample complexity is not established by the proof.
  4. [Lemma C.4, eq. (44)] The inclusion E7 subset A2 used in the proof of local confinement requires the empirical-MLE bound ||theta* - theta_hat*|| <= C7 sqrt(|E|/m) log(|E|/epsilon) to be smaller than the radius in A2. Equation (44), however, only imposes a condition involving C7 sqrt(|E| log|E|)/sqrt(m), with no epsilon-dependent logarithmic factor. As written, the confinement lemma does not follow, and Theorem 3.4 inherits this gap.
minor comments (5)
  1. [Eq. (40)] The definition C40 := (C4 - C3) ∧ (c4 - c3) is negative under Assumption A1, since c4 < c3; it should presumably be (C4 - C3) ∧ (c3 - c4).
  2. [Lemma 4.6 proof] After the inequality a_n - a_{n+1} >= (a_n)^2/(2 L_{i+1} C^2), the constant gamma should be 1/(2 L_{i+1} C^2), not 1/(2 L_{i+1} C).
  3. [Eq. (37) and following display] The Gaussian tail exponent appears off by a factor of 2: with x = t delta/(2 sqrt{|E|} eC19), the bound P(Z >= x) <= exp(-x^2/2) gives exponent -t^2 delta^2/(8 |E| eC19^2), not -t^2 delta^2/(4 |E| eC19^2). This is harmless for the conclusion but should be checked.
  4. [Throughout] There are several typographical errors, including 'specicically', 'lenghts', 'funcitons', 'Hessain', and 'Tehruashvili'. A careful proofreading pass is needed.
  5. [Section 3.4] The O(delta) initialization assumption in Theorem 3.4 is a substantial practical limitation, as the authors acknowledge. A brief discussion of potential initializers, such as distance-based methods that might land in such a box for sufficiently small delta, would strengthen the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the empirical landscape, consistency, and coordinate-descent results are derived from analytic same-author population-landscape lemmas that are independent evidence under the rubric, not from fitted inputs.

full rationale

The derivation chain is not circular under the rubric used here. Theorem 3.2 is proved by applying the uniform matrix Bernstein inequality (Lemma 4.5) to the empirical Hessian, using deterministic Hessian and third-derivative bounds (Lemmas 4.2 and 4.3) and the population landscape result (Theorem 4.4). Theorem 3.3 follows from Theorem 3.2 via Berry-Esseen and a Taylor expansion, and Theorem 3.4 follows from Theorem 3.3 plus the block-coordinate convergence lemma (Lemma 4.6). The population landscape and derivative formulas are taken from prior work by the same authors (Clancy et al. 2025a,b) and are load-bearing, but they are analytic statements with no data-dependent or fitted parameters and with stated assumptions (Assumption A1, delta below the relevant threshold) that do not include the empirical concentration, consistency, or coordinate-descent claims; they are also externally checkable, so under Hard Rule 4 they count as independent evidence rather than circular self-citation. There is no fitted parameter relabeled as a prediction, and no ansatz is smuggled in via citation: the magnetization derivative formulas (13)-(14) are exact identities, not assumptions. The O(delta) initialization requirement is an honest limitation, explicitly flagged by the authors as open (Section 3.4), but it is not a circular step. Separately and non-circularly, the Appendix C proof of Theorem C.1 derives a sufficient sample size (34) with an unabsorbed |E|^3 diam(T) log(1/delta) factor alongside (1/delta)^{diam(T)+8}, so the claimed universal constant C5 in (5) and (31) is not justified as delta tends to 0 and Theorems 3.3 and 3.4 inherit that technical gap; this is a correctness concern, not a circularity, and does not raise the circularity score.

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

The main external input is the population-limit landscape theorem from the same authors' prior preprint (Clancy et al. 2025a), plus derivative formulas proven there. No free parameters are fit to data; the constants C4, C3, c3, c4 and delta define the regime and are not estimated. No new entities are posited.

assumptions (4)
  • domain assumption Population log-likelihood is (C19/delta - 26)-strongly concave and (C19'/delta + 26)-smooth on bTheta0(delta), with theta* the unique maximizer (Thm 4.4, from Clancy et al. 2025a).
    Invoked in the proof of Thm 3.2 to anchor the empirical Hessian eigenvalues via Weyl's inequality; not re-proved in this paper.
  • domain assumption Gradient and Hessian formulas of the single-sample log-likelihood in Lemma 4.2(i,ii) are taken from Clancy et al. 2025a.
    These expressions in terms of magnetizations underlie all Hessian bounds in Lemma 4.3 and the matrix Bernstein verification.
  • ad hoc to paper Assumption A1: true parameters theta* in [1-2C3*delta, 1-2c3*delta]^E and estimated parameters in the larger box [1-2C4*delta, 1-2c4*delta]^E, with C4 > C3 > c3 > c4 and C4 >= 2c4.
    This narrow band of small mutation probabilities is the entire setting of Theorems 3.2 to 3.4; outside it, the likelihood landscape can have multiple global maxima (Steel 1994).
  • domain assumption The initial iterate is within O(delta) of theta* in L2 norm (eq. (40)).
    Needed to keep all coordinate-descent iterates inside the good box bTheta0(delta); the paper does not show how to obtain such a point.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sample Complexity of Branch-length Estimation by Maximum Likelihood." pith.science (2026). https://pith.science/paper/7P4TO7VB

@misc{pith2026250722038,
  author       = {Pith},
  title        = {Pith review of: Sample Complexity of Branch-length Estimation by Maximum Likelihood},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7P4TO7VB}},
  note         = {Machine review of arXiv:2507.22038}
}
abstract

We consider the branch-length estimation problem on a bifurcating tree: a character evolves along the edges of a binary tree according to a two-state symmetric Markov process, and we seek to recover the edge transition probabilities from repeated observations at the leaves. This problem arises in phylogenetics, and is related to latent tree graphical model inference. In general, the log-likelihood function is non-concave and may admit many critical points. Nevertheless, simple coordinate maximization has been known to perform well in practice, defying the complexity of the likelihood landscape. In this work, we provide the first theoretical guarantee as to why this might be the case. We show that deep inside the Kesten-Stigum reconstruction regime, provided with polynomially many $m$ samples (assuming the tree is balanced), there exists a universal parameter regime (independent of the size of the tree) where the log-likelihood function is strongly concave and smooth with high probability. On this high-probability likelihood landscape event, we show that the standard coordinate maximization algorithm converges exponentially fast to the maximum likelihood estimator, which is within $O(1/\sqrt{m})$ from the true parameter, provided a sufficiently close initial point.

Figures

Figures reproduced from arXiv: 2507.22038 by the authors.

Figure 1
Figure 1. A cartoon depiction of non-concave 2D likelihood land￾scape (left) and its contour plot (right). Thm. 3.2 and 3.3 asserts that the empirical likelihood landscape has a well-conditioned strongly concave landscape over a box Θb 0(δ) of size order 1 around the true parameter θ ∗ . Thm. 3.4 asserts that coordinate maximization initialized in the box converges to the MLE θˆ0 in O(1) iterations. and contain numerous local… view at source ↗
Figure 2
Figure 2. A 2D slice of a 5D 2-sample empirical log-likelihood ℓ(θˆ; σ (1), σ(2)) containing θ ∗ on a tree with n = 4 leaves. 1981). The corresponding joint likelihood function in gen￾eral is non-concave, admitting potentially many (complex) critical points (Garc´ıa Puente et al., 2024) for generic data σ (j) . Notably, Steel (1994) provided an explicit example where there are multiple global maximizers to (2) with two sample… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

47 extracted references · 41 canonical work pages

  1. [1]

    More Algorithms for Provable Dictionary Learning

    Arora, S., Bhaskara, A., Ge, R., and Ma, T. More algorithms for provable dictionary learning. arXiv preprint arXiv:1401.0579, 2014

  2. [2]

    Subgradient descent learns orthogonal dictionaries

    Bai, Y., Jiang, Q., and Sun, J. Subgradient descent learns orthogonal dictionaries. In 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, May 6-9, 2019 . OpenReview.net, 2019. URL https://openreview.net/forum?id=HklSf3CqKm

  3. [3]

    Beale, E. M. L., Kendall, M. G., and Mann, D. W. The discarding of variables in multivariate analysis. Biometrika, 54: 0 357--366, 1967. ISSN 0006-3444,1464-3510. doi:10.1093/biomet/54.3-4.357. URL https://doi.org/10.1093/biomet/54.3-4.357

  4. [4]

    and Tetruashvili, L

    Beck, A. and Tetruashvili, L. On the convergence of block coordinate descent type methods. SIAM journal on Optimization, 23 0 (4): 0 2037--2060, 2013

  5. [5]

    M., Ruiz, J., and Zagrebnov, V

    Bleher, P. M., Ruiz, J., and Zagrebnov, V. A. On the purity of the limiting G ibbs state for the I sing model on the B ethe lattice. J. Statist. Phys., 79 0 (1-2): 0 473--482, 1995

  6. [6]

    The K esten- S tigum reconstruction bound is tight for roughly symmetric binary channels

    Borgs, C., Chayes, J., Mossel, E., and Roch, S. The K esten- S tigum reconstruction bound is tight for roughly symmetric binary channels. In 2006 47th Annual IEEE Symposium on Foundations of Computer Science (FOCS'06), pp.\ 518--530. IEEE, 2006

  7. [7]

    Brent, R. P. Algorithms for minimization without derivatives. Courier Corporation, 2013

  8. [8]

    Cavender, J. A. Taxonomy with confidence. Mathematical biosciences, 40 0 (3-4): 0 271--280, 1978

Show all 47 references
  1. [9]

    and Chen, Y

    Chen, X. and Chen, Y. Inference for high-dimensional sparse econometric models. Annual Review of Economics, 11: 0 291--317, 2019

  2. [10]

    M., and Chen, Y

    Chi, Y., Lu, Y. M., and Chen, Y. Nonconvex optimization meets low-rank matrix factorization: An overview. Trans. Sig. Proc., 67 0 (20): 0 5239–5269, October 2019. ISSN 1053-587X. doi:10.1109/TSP.2019.2937282. URL https://doi.org/10.1109/TSP.2019.2937282

  3. [11]

    and Tuller, T

    Chor, B. and Tuller, T. Finding a maximum likelihood tree is hard. Journal of the ACM (JACM), 53 0 (5): 0 722--744, 2006

  4. [12]

    , D., Lyu, H., and Roch, S

    Clancy, Jr. , D., Lyu, H., and Roch, S. Likelihood landscape of binary latent model on a tree. arXiv:2501.17622, 2025 a

  5. [13]

    , D., Lyu, H., Roch, S., and Sly, A

    Clancy, Jr. , D., Lyu, H., Roch, S., and Sly, A. Likelihood-based root state reconstruction on a tree: Robustness to parameters. arXiv:2501.13208, 2025 b

  6. [14]

    Mathematical M ethods of S tatistics , volume 9 of Princeton Mathematical Series

    Cram\'er, H. Mathematical M ethods of S tatistics , volume 9 of Princeton Mathematical Series. Princeton University Press, Princeton, NJ, 1946

  7. [15]

    Evolutionary trees and the ising model on the bethe lattice: a proof of steel's conjecture

    Daskalakis, C., Mossel, E., and Roch, S. Evolutionary trees and the ising model on the bethe lattice: a proof of steel's conjecture. Probability Theory and Related Fields, 149: 0 149--189, 2011. ISSN 0178-8051. URL http://dx.doi.org/10.1007/s00440-009-0246-2. 10.1007/s00440-009-0246-2

  8. [16]

    Probability: Theory and Examples , volume 49

    Durrett, R. Probability: Theory and Examples , volume 49. Cambridge University Press , 2019

  9. [17]

    Farris, J. S. A probability model for inferring evolutionary trees. Systematic Biology, 22 0 (3): 0 250--256, 1973

  10. [18]

    Evolutionary trees from DNA sequences: A maximum likelihood approach

    Felsenstein, J. Evolutionary trees from DNA sequences: A maximum likelihood approach. Journal of Molecular Evolution, 17 0 (6): 0 368--376, 1981

  11. [19]

    and Tateno, Y

    Fukami, K. and Tateno, Y. On the maximum likelihood method for estimating molecular trees: uniqueness of the likelihood point. Journal of Molecular Evolution, 28: 0 460--464, 1989

  12. [20]

    D., Garrote-L \'o pez, M., and Shehu, E

    Garc \' a Puente, L. D., Garrote-L \'o pez, M., and Shehu, E. Computing algebraic degrees of phylogenetic varieties. Algebraic Statistics, 14 0 (2): 0 215--231, 2024

  13. [21]

    and Gascuel, O

    Guindon, S. and Gascuel, O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic biology, 52 0 (5): 0 696--704, 2003

  14. [22]

    New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0

    Guindon, S., Dufayard, J.-F., Lefort, V., Anisimova, M., Hordijk, W., and Gascuel, O. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Systematic Biology, 59 0 (3): 0 307--321, 2010

  15. [23]

    Hocking, R. R. and Leslie, R. N. Selection of the best subset in regression analysis. Technometrics, 9: 0 531--540, 1967. ISSN 0040-1706,1537-2723. doi:10.2307/1266192. URL https://doi.org/10.2307/1266192

  16. [24]

    On the extremality of the disordered state for the I sing model on the B ethe lattice

    Ioffe, D. On the extremality of the disordered state for the I sing model on the B ethe lattice. Lett. Math. Phys., 37 0 (2): 0 137--143, 1996

  17. [25]

    Non-convex optimization for machine learning

    Jain, P., Kar, P., et al. Non-convex optimization for machine learning. Foundations and Trends in Machine Learning , 10 0 (3-4): 0 142--363, 2017

  18. [26]

    Linear convergence of gradient and proximal-gradient methods under the P olyak- ojasiewicz condition

    Karimi, H., Nutini, J., and Schmidt, M. Linear convergence of gradient and proximal-gradient methods under the P olyak- ojasiewicz condition. In Frasconi, P., Landwehr, N., Manco, G., and Vreeken, J. (eds.), Machine Learning and Knowledge Discovery in Databases, pp.\ 795--811,...

  19. [27]

    and Stigum, B

    Kesten, H. and Stigum, B. P. Additional limit theorems for indecomposable multidimensional galton-watson processes. The Annals of Mathematical Statistics, 37 0 (6): 0 1463--1481, 1966

  20. [28]

    and Babu, G

    Li, B. and Babu, G. J. A Graduate Course on Statistical Inference . Springer Texts in Statistics. Springer, New York, 2019. ISBN 978-1-4939-9759-6; 978-1-4939-9761-9. doi:10.1007/978-1-4939-9761-9. URL https://doi.org/10.1007/978-1-4939-9761-9

  21. [29]

    A topological property of real analytic subsets

    Lojasiewicz, S. A topological property of real analytic subsets. Coll. du CNRS, Les \'e quations aux d \'e riv \'e es partielles , 117 0 (87-89): 0 2, 1963

  22. [30]

    Implicit regularization in nonconvex statistical estimation: G radient descent converges linearly for phase retrieval, matrix completion, and blind deconvolution

    Ma, C., Wang, K., Chi, Y., and Chen, Y. Implicit regularization in nonconvex statistical estimation: G radient descent converges linearly for phase retrieval, matrix completion, and blind deconvolution. Found. Comput. Math., 20: 0 451--632, 2020

  23. [31]

    Combinatorial statistics and the sciences

    Mossel, E. Combinatorial statistics and the sciences. In Proceedings of the ICM, 2022

  24. [32]

    Murphy, K. P. Machine Learning: A Probabilistic Perspective . MIT Press, 2012

  25. [33]

    Molecular studies of evolution: a source of novel statistical problems

    Neyman, J. Molecular studies of evolution: a source of novel statistical problems. In Statistical decision theory and related topics, pp.\ 1--27. Elsevier, 1971

  26. [34]

    A., von Haeseler, A., and Minh, B

    Nguyen, L.-T., Schmidt, H. A., von Haeseler, A., and Minh, B. Q. IQ-TREE : a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Molecular Biology and Evolution, 32 0 (1): 0 268--274, 2015

  27. [35]

    Polyak, B. T. Gradient methods for the minimisation of functionals. USSR Computational Mathematics and Mathematical Physics, 3 0 (4): 0 864--878, 1963

  28. [36]

    A short proof that phylogenetic tree reconstruction by maximum likelihood is hard

    Roch, S. A short proof that phylogenetic tree reconstruction by maximum likelihood is hard. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 3 0 (1): 0 92--94, 2006

  29. [37]

    Modern Discrete Probability: An Essential Toolkit

    Roch, S. Modern Discrete Probability: An Essential Toolkit. Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, 2024. doi:10.1017/9781009305129

  30. [38]

    and Sly, A

    Roch, S. and Sly, A. Phase transition in the sample complexity of likelihood-based phylogeny inference. Probability Theory and Related Fields, 169 0 (1): 0 3--62, Oct 2017. doi:10.1007/s00440-017-0793-x. URL https://doi.org/10.1007/s00440-017-0793-x

  31. [39]

    RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies

    Stamatakis, A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30 0 (9): 0 1312--1313, 2014

  32. [40]

    The maximum likelihood point for a phylogenetic tree is not unique

    Steel, M. The maximum likelihood point for a phylogenetic tree is not unique. Systematic Biology, 43 0 (4): 0 560--564, 1994

  33. [41]

    Phylogeny: Discrete and Random Processes in Evolution

    Steel, M. Phylogeny: Discrete and Random Processes in Evolution . CBMS - NSF Regional Conference Series in Applied Mathematics . Society for Industrial and Applied Mathematics, 2016. ISBN 978-1-61197-447-8. URL https://books.google.com/books?id=tddCDQAAQBAJ

  34. [42]

    A geometric analysis of phase retrieval

    Sun, J., Qu, Q., and Wright, J. A geometric analysis of phase retrieval. Found. Comput. Math., 18 0 (5): 0 1131--1198, 2018. ISSN 1615-3375,1615-3383. doi:10.1007/s10208-017-9365-9. URL https://doi.org/10.1007/s10208-017-9365-9

  35. [43]

    Tropp, J. A. User-friendly tail bounds for sums of random matrices. Found. Comput. Math., 12: 0 389--434, 2012

  36. [44]

    Note on the consistency of the maximum likelihood estimate

    Wald, A. Note on the consistency of the maximum likelihood estimate. The Annals of Mathematical Statistics, 20 0 (4): 0 595--601, 1949

  37. [45]

    Computational Phylogenetics : An Introduction to Designing Methods for Phylogeny Estimation

    Warnow, T. Computational Phylogenetics : An Introduction to Designing Methods for Phylogeny Estimation . Cambridge University Press, 2018. ISBN 978-1-107-18471-8. URL https://books.google.com/books?id=7-k4DwAAQBAJ

  38. [46]

    From symmetry to geometry: Tractable nonconvex problems

    Zhang, Y., Qu, Q., and Wright, J. From symmetry to geometry: Tractable nonconvex problems. arXiv preprint arXiv:2007.06753, 2020

  39. [47]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.