Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

Pathwise optimization for bridge-type estimators and its applications

T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read The full solution path of a nonconvex multi-penalty bridge estimator can be computed directly by proximal thresholding updates, without convex relaxation, and the resulting path estimates are pointwise consistent.

desk verdict Computationally plausible and worth a referee, but the q<1 path-consistency theorem assumes the very closeness it needs to prove. read the letter →

arxiv 2412.04047 v1 pith:DE7R27L2 submitted 2024-12-05 stat.ML cs.LGmath.STstat.COstat.TH

classification stat.MLcs.LGmath.STstat.COstat.TH MSC 62J0762F1265K1090C2662M05
keywords bridge-typeestimatorsadaptivethresholdingoperatorsnonconvexoptimizationpathconsistencyproximalmapsregularizedLqpenaltiesstochasticdifferentialequations
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

Bridge-type estimators minimize a loss plus a sum of $\ell^q$ penalties with $q\in(0,1]$; for $q<1$ the objective is nonconvex and nondifferentiable, which is why standard practice replaces the penalty by a convex surrogate. This paper claims that the exact solution path of such estimators can nevertheless be computed, by iterating adaptive thresholding operators that are proximal maps of the $\ell^q$ penalty. Two algorithms are proposed: a monitored accelerated proximal gradient method and a blockwise proximal alternating (PALM) scheme; both are shown to converge, for each $\lambda$ and sample size, to a critical point of the bridge objective. The paper further claims path consistency: after normalization by the estimator's convergence rates, the computed path at any fixed $\lambda$ is bounded in probability, with oracle selection properties under additional assumptions. A reader should care because exact computation, if correct, lets practitioners use bridge estimators' oracle behavior directly rather than through convex relaxations.

What carries the argument

The load-bearing object is the adaptive thresholding operator $T^q_{\lambda w}(z) = (T^q_{\lambda w_1}(z_1),\dots,T^q_{\lambda w_p}(z_p))^\top$, defined as the proximal map of the weighted $\ell^q$ penalty, whose one-dimensional form generalizes soft thresholding (for $q=1$) and hard thresholding (as $q\to 0$). Every iteration of both algorithms is a gradient step followed by this operator, so the nonconvex penalty is used exactly, not relaxed. The convergence arguments rely on the monitored acceleration scheme for nonconvex proximal gradient methods and on the Kurdyka--Lojasiewicz/semi-algebraic framework for PALM, which together ensure that accumulation points are critical points; the path-consistency argument additionally uses the formula for $\lambda_{\max}$, the smallest penalty at which the null vector is stationary.

What would settle it

Run Algorithm 2 on the diffusion model of Section 7.2 from a starting point far from the quasi-likelihood estimate, at a fixed $\lambda$, and record whether the iterates converge to the same critical point as when the algorithm starts near the estimate; if the final points differ, or if the normalized path error $A_n^{-1}(\hat{\theta}_n(\lambda)-\theta_0)$ fails to stay bounded across $n$ for such starts, then the basin assumption required by Theorem 1(ii) is not satisfied in that setting.

Watch

Extended reading notes

Core claim

The central discovery is that the bridge-type estimation problem (3.5), despite its nonconvexity, admits a pathwise optimization scheme whose updates are componentwise applications of the adaptive $q$-thresholding operator $T^q_{\lambda w}$, the proximal map of the weighted $\ell^q$ penalty. Algorithm 1, a monitored accelerated proximal gradient method, and Algorithm 2, a blockwise proximal alternating linearized minimization method, both generate bounded sequences that converge to a critical point with $0 \in \partial G_n(\theta^*; q)$. Theorem 1 states that, under the mixed-rates assumptions (3.1), (5.6), and (5.7), the path estimates satisfy $A_n^{-1}(\hat{\theta}_n(\lambda) - \theta_0)1_{\{\lambda_{\max}(n)>\lambda\}} = O_p(1)$ for the LASSO case for either algorithm, and for the bridge case $q<1$ for Algorithm 2 provided it is started sufficiently close to the bridge-type estimate; under further assumptions the oracle properties of selection and asymptotic normality transfer to the path.

Load-bearing premise

For $q<1$, the proof that the computed path is consistent assumes the algorithm is started very close to the bridge-type estimator it is supposed to recover, and the paper does not provide a check or guarantee for that starting condition.

Editorial extensions

If this is right

  • For every $\lambda$ and sample size $n$, the path output by either algorithm converges to a critical point of the bridge objective, so the computed path is a genuine stationary-point path rather than a relaxed approximation.
  • Because the thresholding operator is evaluated componentwise, the algorithms scale to high-dimensional parameter vectors and to multiple penalty groups with different exponents $q_i$ and different convergence rates.
  • When $q=1$, the path estimates are pointwise consistent for either algorithm; when $q<1$, the same consistency holds for the PALM algorithm under a basin-of-attraction condition on the starting point.
  • In the diffusion-process application, the bridge path outperforms the LASSO path in selection probability and is less sensitive to suboptimal tuning, and the blockwise scheme reduces iteration counts by roughly a factor of three relative to coordinate descent.
  • Under the additional assumptions imported from the authors' earlier work, the path estimates inherit the oracle properties of variable selection and asymptotic normality.

Reading between the lines

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

  • One extension the paper leaves implicit: warm-starting each $\lambda$ from the previous solution, recommended in the paper for speed, may in practice place the algorithm inside the basin required by Theorem 1(ii); this could be tested by comparing warm starts against random starts on the same problem.
  • The discontinuity of the path at the threshold points suggests a data-driven rule for choosing $\lambda$, such as selecting the largest $\lambda$ before a variable drops out, which the paper does not explore.
  • The same proximal-thresholding machinery applies to other concave penalties with known proximal maps, such as SCAD or MCP, offering a route to exact pathwise computation for those estimators as well.
  • A practical consequence of the basin assumption is that path consistency for $q<1$ is only guaranteed for favorable initializations; checking it empirically with multiple random starts would clarify how often the condition holds in real data.
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 / 4 minor

Summary. The paper proposes two pathwise algorithms for computing adaptive bridge-type estimators with multiple nonconvex penalties of the form sum of weighted l_q norms, q in (0,1]. Algorithm 1 is a monotone accelerated proximal gradient method that uses the exact adaptive thresholding operator T^q for the q-norm proximal map; Algorithm 2 is a blockwise proximal alternating linearized minimization (PALM) method. The authors claim convergence to critical points for both algorithms, derive a formula for the largest penalty lambda_max, state a path-consistency theorem (Theorem 1), and demonstrate the methods on GLM regression and on discretely observed ergodic diffusion processes. The computational machinery is mostly imported from existing nonconvex optimization theory, while the statistical oracle properties are imported from the authors' earlier paper [16].

Significance. If fully established, the paper would be a useful contribution because it computes the bridge path directly from the nonconvex, nonsmooth objective rather than through a convex relaxation, and it handles mixed-rate asymptotics via blockwise penalties. The thresholding formulas and the application to diffusion processes are concrete strengths, and the convergence-to-critical-point claims for the two algorithms are plausible from the cited KL/semi-algebraic framework. The main new statistical claim, however, is path consistency for 0<q_i<1 in Theorem 1(ii), and that claim is not supported as stated because it assumes an unverified basin-of-attraction condition. The LASSO case (Theorem 1(i)) is more defensible, but it also relies on oracle properties imported from [16] rather than being proved here.

major comments (4)
  1. [Section 4.2, property 2; Section 5, Theorem 1(ii); Remark 5.1] The path-consistency result for 0<q_i<1 is conditional on the starting point being 'quite close' to the bridge-type estimate, with Theorem 2.12 of [3] cited for local convergence. The pathwise protocol in Remark 5.1 initializes at the zero vector at lambda_max and then warm-starts from the previous lambda-solution; the manuscript contains no argument that such starts lie in the basin of attraction of the global bridge estimator for large n, nor any verifiable condition that a user can check. As stated, Theorem 1(ii) assumes essentially that the algorithm already converges to the estimator whose consistency is being claimed, making the q<1 path-consistency claim circular for the actual pathwise outputs.
  2. [Section 5, proof of Theorem 1(ii)] The proof of part (ii) consists of the sentence 'In the bridge case the argument is analogous, where (local) convergence to a global minimum is guaranteed under the same conditions discussed for PALM Algorithm.' This does not establish the theorem. For q<1, Algorithm 2 is only guaranteed to converge to a critical point (Theorem 1 in [8]), and local convergence to the global minimizer is exactly the basin-of-attraction condition flagged above. The proof needs to show that, under the prescribed pathwise initialization, the output of Algorithm 2 satisfies A_n^{-1}(hat_theta_n(lambda)-theta_0) 1{lambda_max^{(n)}>lambda}=O_p(1); no such argument is supplied.
  3. [Section 4.1, Eq. (4.4)] The extrapolation step in Algorithm 1 uses c_{t-1} and zeta^t, but the initialization only sets c_1=1 and theta^1=theta^0=eta_0; c_0 and zeta^1 are never defined, so for t=1 the quantities c_{t-1}/c_t and zeta^t are undefined. The update should be rewritten in the standard FISTA/monotone-APG form of [29], or explicit initialization of c_0 and zeta^0 should be provided. As written, Algorithm 1 is not reproducible from the pseudocode.
  4. [Section 5, Eq. (5.1) and opening sentence] The paper defines lambda_max using the phrase 'stationary point' but then characterizes it by the fixed-point condition T^q_{lambda s w}(0)=0. For 0<q<1 the thresholding operator is discontinuous, and the two notions are not equivalent: the fixed-point condition is sufficient for the algorithm to output zero but does not, by itself, imply the subdifferential stationarity condition 0 in partial G_n(0;q) used in (4.9). Since Lemma 1 and Theorem 1 use lambda_max^{(n)} to define the non-null part of the path, this distinction needs to be clarified and proved.
minor comments (4)
  1. [Lemma 1, display after Eq. (5.9)] The constant is written c_{q_i}^{-2}, whereas Proposition 1 and Eq. (5.2) require c_{q_i}^{-(2-q_i)}; please correct the exponent.
  2. [Section 7.2 and Figure 6] The text and the Figure 6 caption refer to 'Algorithms 1 and 3' and 'PALM (Alg. 3)', but only Algorithms 1 and 2 are defined in the paper; the references should be to Algorithm 2.
  3. [Section 7.2, Table 2] The number of Monte Carlo replications used to estimate MSE and P0 is not reported; please state the number of simulated datasets.
  4. [Throughout] There are several typos and informal phrases: 'adpative' in the abstract, 'shed in light' in Section 3, and 'yi in R^n' in Section 6.2 where y in R^n is intended.

Circularity Check

1 steps flagged · score 6.0 of 10

The q<1 path-consistency theorem assumes a starting point in the basin of the bridge estimator it is supposed to compute; Theorem 1(ii) is therefore conditional on the very object the path algorithm is meant to produce.

  1. self definitional [Section 4.2, property 2; Section 5, Theorem 1(ii) and its proof; Remark 5.1]
    "2) if we chose a starting point quite close to the bridge-type estimate ˇθn, then {ˇθt : t = 1, 2, ...} tends to ˇθn (see Theorem 2.12 in [3] for the a detailed statement). ... (ii) the block-wise Algorithm 2, for 0 < qi < 1, if the starting point is quite close to ˆθn (see Theorem 2.12 in [3]); are pointwise consistent ... In the bridge case the argument is analogous, where (local) convergence to a global minimum is guaranteed under the same conditions discussed for PALM Algorithm."

    Theorem 1(ii) promises path consistency for the estimates produced by Algorithm 2 only under the condition that the algorithm is started 'quite close to ˆθn', where ˆθn is the bridge-type estimator whose consistency is the content of the claim. Section 4.2 property 2 is the only mechanism guaranteeing convergence to ˆθn, and it is exactly this basin-of-attraction premise. The actual pathwise initialization in Remark 5.1 starts at the zero vector at λmax and warm-starts from the previous λ-solution; nothing in the paper verifies that these starts lie in the basin of ˆθn. The proof of (ii) simply repeats the PALM condition ('the argument is analogous') rather than deriving it from the algorithm's initialization.

full rationale

The algorithmic convergence parts of the paper are not circular: the convergence of Algorithm 1 to critical points (4.9) is grounded in the external KL/semi-algebraic framework of [29], and the PALM convergence of Algorithm 2 is grounded in [8]. The λmax computation and its divergence in Lemma 1 are derived from the thresholding operators and the stated statistical assumptions, not from the target result. The circular burden is concentrated in the q<1 path-consistency claim, Theorem 1(ii). There, the paper's own sufficient condition is that the algorithm be started close to ˆθn, the very bridge-type estimator the path is supposed to compute. Since the prescribed pathwise initialization is not shown to satisfy this condition, the theorem does not establish path consistency for the actual algorithm; it states a conditional implication whose premise already contains the conclusion. The additional oracle properties imported from [16] are prior-work statistical properties of the bridge estimator and are not by themselves circular, but they do not repair the missing basin-of-attraction guarantee. Overall, the central optimization contribution retains independent content, so the score reflects partial circularity in the statistical path-consistency theorem rather than full equivalence.

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

The paper introduces no new physical or mathematical entities; it imports existing thresholding operators, convergence theorems, and the multi-penalty bridge framework from prior work, including the authors' own [16]. The main extra loads are the asymptotic regularity assumptions on the loss and estimator, the semi-algebraic condition on the penalties, and the basin-of-attraction condition for the nonconvex path-consistency result.

free parameters (4)
  • Penalty exponents q_i = q1=q2=1/2 in diffusion simulation; q not specified in theory
    The user selects q_i; the paper uses 1/2 in the SDE simulation study. The theory allows any q_i in (0,1].
  • Adaptive weight exponents δ_i = δ1=δ2=4 in diffusion simulation
    The simulation chooses weights w = 1/|initial estimator|^4. The theory requires δ_i > 1 - q_i, and constants work for δ_i > 2 - q_i, so the simulation value is a user choice.
  • Algorithm step-size constants s, u, α_i = Not reported numerically; s,u < 1/L, 0 < α_i < 1
    These are algorithmic hyperparameters needed to define the updates in Algorithm 1 and Algorithm 2. They are not fitted to data but are chosen by the practitioner.
  • Penalization parameter λ = Selected by cross-validation; reported λopt ≈ 0.5 in diffusion study
    λ is the path variable itself. In simulations it is chosen by cross-validation, which is standard tuning rather than a free parameter in the theoretical derivation.
assumptions (6)
  • domain assumption The loss L_n is proper, coercive, C^2 with Lipschitz-continuous gradient, and level sets {G_n ≤ G_n(θ0)} are contained in int(Θ).
    Assumed in Section 4 before the algorithms and needed to apply the convergence results of [29] and [8].
  • domain assumption The unpenalized estimator \tilde θ_n satisfies mixed-rates consistency and asymptotic normality as in equation (3.1).
    This asymptotic framework from [35] underlies the multi-penalty structure and the oracle-property results invoked in Section 5.
  • domain assumption D_n = A_n G_n A_n converges in probability to a positive definite D and a_i^n r_i^n = O_p(1), as in assumptions (5.6) and (5.7).
    These are needed for Lemma 1 and for the oracle properties of the bridge estimator from [16].
  • ad hoc to paper The bridge-type estimator from [16] has selection consistency and asymptotic normality under additional technical assumptions.
    Theorem 1 delegates the oracle properties of the updates to the authors' own preceding paper [16], and the path-consistency conclusion depends on this imported result.
  • standard math For Algorithm 2, the penalty exponents q_i lie in (0,1] ∩ Q so that \|·\|_{q_i}^{q_i} is semi-algebraic and G_n is a KL function.
    Stated in Section 4.2; the semi-algebraic property is needed to apply Theorem 1 in [8] and the convergence results for PALM.
  • domain assumption If the algorithm starts 'quite close' to the bridge-type estimate, the PALM sequence converges to that estimator, as in Theorem 2.12 of [3].
    This locality condition is the weakest assumption: it is not checked or constructed by the algorithm, and Theorem 1 (ii) depends on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pathwise optimization for bridge-type estimators and its applications." pith.science (2026). https://pith.science/paper/DE7R27L2

@misc{pith2026241204047,
  author       = {Pith},
  title        = {Pith review of: Pathwise optimization for bridge-type estimators and its applications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DE7R27L2}},
  note         = {Machine review of arXiv:2412.04047}
}
abstract

Sparse parametric models are of great interest in statistical learning and are often analyzed by means of regularized estimators. Pathwise methods allow to efficiently compute the full solution path for penalized estimators, for any possible value of the penalization parameter $\lambda$. In this paper we deal with the pathwise optimization for bridge-type problems; i.e. we are interested in the minimization of a loss function, such as negative log-likelihood or residual sum of squares, plus the sum of $\ell^q$ norms with $q\in(0,1]$ involving adpative coefficients. For some loss functions this regularization achieves asymptotically the oracle properties (such as the selection consistency). Nevertheless, since the objective function involves nonconvex and nondifferentiable terms, the minimization problem is computationally challenging. The aim of this paper is to apply some general algorithms, arising from nonconvex optimization theory, to compute efficiently the path solutions for the adaptive bridge estimator with multiple penalties. In particular, we take into account two different approaches: accelerated proximal gradient descent and blockwise alternating optimization. The convergence and the path consistency of these algorithms are discussed. In order to assess our methods, we apply these algorithms to the penalized estimation of diffusion processes observed at discrete times. This latter represents a recent research topic in the field of statistics for time-dependent data.

Figures

Figures reproduced from arXiv: 2412.04047 by the authors.

Figure 2
Figure 2. Comparison between LASSO and Bridge path. particular features of the estimator, which are variable selection combined with non￾continuous estimation path and penalization vanishing for large values of lambda [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. Comparison of test errors as a function of normalized penal￾ization parameter λ/λmax for bridge (solid line), LLA (dashed), LASSO (dotted). Figure B is a zoom of Figure A around the minimum of the curves. lower test error for suitable values of the tuning parameter, at the cost of a more rapid increase in bias as the penalization parameter increases. 7. Application to Stochastic Differential Equations 7.1. Ergodic d… view at source ↗
Figure 4
Figure 4. A simulated sample path of the solution to (7.4) where A = (αi,j : i, j = 1, . . . d) and B = (βi,j : i, j = 1, . . . d) are parameter matrices assumed to be positive definite. In our case d = 4 and the true parameter matrices are (7.5) A =     4 −1.8 0 0 0 4 −1.8 0 0 0 4 −1.8 0 0 0 4     B =     4 0 0 0 0 4 0 0 0 0 4 0 0 0 0 4     In this case the process components are not independent because of th… view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Model selection proportions for n = 103 as a function of the normalized penalization parameter. 0.0 0.25 0.5 0.75 1.0 log λ/ log λmax 5 10 15 20 25 30 35 APGD PALM CD lasso bridge [PITH_FULL_IMAGE:figures/full_fig_p023_5.png]
Figure 6
Figure 6. Figure 6: Comparison between algorithms APGD (Alg.1), PALM (Alg. 3), CD (Coordinate Descent) for lasso (dashed) and bridge (solid line): number of iterations as a function of the normalized penalization param￾eter. In CD in the bridge case is meant to be (Alg.3) with one-paramet…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Adaptive Elastic-Net estimation for sparse diffusion processes

    math.ST 2024-12 conditional novelty 6.0 of 10

    Adaptive Elastic-Net for ergodic diffusions achieves mixed-rate oracle properties and non-asymptotic l2 and prediction error bounds.

  2. Ergodic Network Stochastic Differential Equations

    stat.ME 2024-12 reject novelty 5.0 of 10

    A quasi-likelihood and adaptive Lasso framework is proposed for estimating ergodic network SDE models with nonlinear drift, stochastic volatility, and directed graphs.

Reference graph

Works this paper leans on

49 extracted references · 43 canonical work pages · cited by 2 Pith papers

  1. [29]

    E. L. Lehmann, J. P. Romano, and G. Casella. Testing statistical hypotheses , volume 3. Springer, 2005

  2. [8]

    Bolte, S

    J. Bolte, S. Sabach, and M. Teboulle. Proximal alternating linearized minimization for nonconvex and nonsmooth problems. Mathematical Programming, 146(1):459–494, 2014

  3. [16]

    De Gregorio and S

    A. De Gregorio and S. M. Iacus. Adaptive lasso-type estimation for multivariate diffusion processes. Econometric Theory, 28(4):838–860, 2012

  4. [3]

    Attouch, J

    H. Attouch, J. Bolte, and B. F. Svaiter. Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward–backward splitting, and regularized gauss–seidel methods. Mathematical Programming, 137:91–129, 2013

  5. [1]

    Antoine and E

    B. Antoine and E. Renault. Efficient minimum distance estimation with multiple rates of conver- gence. Journal of Econometrics , 170(2):350–367, 2012

  6. [2]

    Attouch, J

    H. Attouch, J. Bolte, P. Redont, and A. Soubeyran. Proximal alternating minimization and projec- tion methods for nonconvex problems: An approach based on the kurdyka- Lojasiewicz inequality. Mathematics of Operations Research, 35(2):438–457, 2010

  7. [4]

    Beck and M

    A. Beck and M. Teboulle. Fast gradient-based algorithms for constrained total variation image denoising and deblurring problems. IEEE Transactions on Image Processing , 18(11):2419–2434, 2009

  8. [5]

    Beck and M

    A. Beck and M. Teboulle. A fast iterative shrinkage-thresholding algorithm for linear inverse prob- lems. SIAM Journal on Imaging Sciences , 2(1):183–202, 2009

Show all 49 references
  1. [6]

    Bertsekas

    D. Bertsekas. Convex Optimization Theory . Athena Scientific optimization and computation series. Athena Scientific, 2009

  2. [7]

    Bertsekas

    D. Bertsekas. Convex Optimization Algorithms . Athena Scientific, 2015

  3. [9]

    Boyd and L

    S. Boyd and L. Vandenberghe. Convex Optimization . Cambridge University Press, 2004

  4. [10]

    Breheny and J

    P. Breheny and J. Huang. Coordinate descent algorithms for nonconvex penalized regression, with applications to biological feature selection. The annals of applied statistics , 5(1):232, 2011

  5. [11]

    E. J. Candes, M. B. Wakin, and S. P. Boyd. Enhancing sparsity by reweighted ℓ1 minimization. Journal of Fourier analysis and applications , 14(5):877–905, 2008

  6. [12]

    F. Chen, L. Shen, and B. Suter. Computing the proximity operator of the ℓp norm with 0 < p <

  7. [13]

    Publisher Copyright: © The Institution of Engineering and Technology 2016

    IET Signal Processing , 10(5):557–565, July 2016. Publisher Copyright: © The Institution of Engineering and Technology 2016

  8. [14]

    Cio lek, D

    G. Cio lek, D. Marushkevych, and M. Podolskij. On dantzig and lasso estimators of the drift in a high dimensional ornstein-uhlenbeck model. Electronic Journal of Statistics, 14(2):4395–4420, 2020. 26 ALESSANDRO DE GREGORIO AND FRANCESCO IAFRATE

  9. [15]

    Ciolek, D

    G. Ciolek, D. Marushkevych, and M. Podolskij. On lasso estimator for the drift function in diffusion models. arXiv preprint arXiv:2209.05974 , 2022

  10. [17]

    De Gregorio and F

    A. De Gregorio and F. Iafrate. Regularized bridge-type estimation with multiple penalties. Annals of the Institute of Statistical Mathematics , pages 1–31, 2020

  11. [18]

    Fan and R

    J. Fan and R. Li. Variable selection via nonconcave penalized likelihood and its oracle properties. Journal of the American Statistical Association , 96(456):1348–1360, 2001

  12. [19]

    L. E. Frank and J. H. Friedman. A statistical view of some chemometrics regression tools. Techno- metrics, 35(2):109–135, 1993

  13. [20]

    Fu and K

    W. Fu and K. Knight. Asymptotics for lasso-type estimators. The Annals of statistics , 28(5):1356– 1378, 2000

  14. [21]

    Hastie, R

    T. Hastie, R. Tibshirani, and M. Wainwright. Statistical Learning with Sparsity: The Lasso and Generalizations. Chapman & Hall/CRC, 2015

  15. [22]

    Huang, J

    J. Huang, J. L. Horowitz, and S. Ma. Asymptotic properties of bridge estimators in sparse high- dimensional regression models. The Annals of Statistics , 36(2):587–613, 2008

  16. [23]

    Huang, S

    J. Huang, S. Ma, H. Xie, and C.-H. Zhang. A group bridge approach for variable selection. Biometrika, 96(2):339–355, 2009

  17. [24]

    D. R. Hunter and R. Li. Variable selection using mm algorithms. Annals of statistics , 33(4):1617, 2005

  18. [25]

    Kamatani and M

    K. Kamatani and M. Uchida. Hybrid multi-step estimators for stochastic differential equations based on sampled data. Statistical Inference for Stochastic Processes, 18(2):177–204, 2015

  19. [26]

    M. Kessler. Estimation of an ergodic diffusion from discrete observations. Scandinavian Journal of Statistics, 24(2):211–229, 1997

  20. [27]

    Kinoshita and N

    Y. Kinoshita and N. Yoshida. Penalized quasi likelihood estimation for variable selection. arXiv preprint arXiv:1910.12871, 2019

  21. [28]

    Y. Koike. De-biased graphical lasso for high-frequency data. Entropy, 22(4), 2020

  22. [30]

    Li and Z

    H. Li and Z. Lin. Accelerated proximal gradient methods for nonconvex programming. Advances in neural information processing systems , 28:379–387, 2015

  23. [31]

    Marjanovic and V

    G. Marjanovic and V. Solo. On lq optimization and matrix completion. IEEE Transactions on Signal Processing, 60(11):5714–5724, 2012

  24. [32]

    Masuda and Y

    H. Masuda and Y. Shimizu. Moment convergence in regularized estimation under multiple and mixed-rates asymptotics. Mathematical Methods of Statistics , 26(2):81–110, 2017

  25. [33]

    Mazumder, J

    R. Mazumder, J. H. Friedman, and T. Hastie. Sparsenet: Coordinate descent with nonconvex penalties. Journal of the American Statistical Association , 106(495):1125–1138, 2011

  26. [34]

    Y. E. Nesterov. A method for unconstrained convex minimization problem with the rate of conver- gence o(1/k2). Soviet Mathematics Doklady , 27(2):372–376, 1983

  27. [35]

    Pock and S

    T. Pock and S. Sabach. Inertial proximal alternating linearized minimization (ipalm) for nonconvex and nonsmooth problems. SIAM Journal on Imaging Sciences , 9(4):1756–1787, 2016

  28. [36]

    Radchenko

    P. Radchenko. Mixed-rates asymptotics. The Annals of Statistics , 36(1):287–309, 2008

  29. [37]

    Rockafellar and R

    R. Rockafellar and R. J.-B. Wets. Variational Analysis. Springer Verlag, Heidelberg, Berlin, New York, 2009

  30. [38]

    Suzuki and N

    T. Suzuki and N. Yoshida. Penalized least squares approximation methods and their applications to stochastic processes. Japanese Journal of Statistics and Data Science , 3, 12 2020

  31. [39]

    Tibshirani

    R. Tibshirani. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society. Series B (Methodological) , 58(1):267–288, 1996

  32. [40]

    P. Tseng. Convergence of a block coordinate descent method for nondifferentiable minimization. Journal of optimization theory and applications , 109(3):475–494, 2001

  33. [41]

    Uchida and N

    M. Uchida and N. Yoshida. Adaptive estimation of an ergodic diffusion process based on sampled data. Stochastic Processes and their Applications , 122(8):2885–2924, 2012. PATHWISE OPTIMIZATION FOR BRIDGE-TYPE ESTIMATORS AND ITS APPLICATIONS 27

  34. [42]

    Uchida and N

    M. Uchida and N. Yoshida. Adaptive bayes type estimators of ergodic diffusion processes from discrete observations. Statistical Inference for Stochastic Processes, 17(2):181–219, 2014

  35. [43]

    Wang and C

    H. Wang and C. Leng. Unified lasso estimation by least squares approximation. Journal of the American Statistical Association, 102(479):1039–1048, 2007

  36. [44]

    Xu and W

    Y. Xu and W. Yin. A globally convergent algorithm for nonconvex optimization based on block coordinate update. Journal of Scientific Computing , 72(2):700–734, 2017

  37. [45]

    Z. Xu, X. Chang, F. Xu, and H. Zhang. l1/2 regularization: A thresholding representation theory and a fast solver. IEEE Transactions on Neural Networks and Learning Systems , 23(7):1013–1027, 2012

  38. [46]

    N. Yoshida. Polynomial type large deviation inequalities and quasi-likelihood analysis for stochastic differential equations. Annals of the Institute of Statistical Mathematics , 63(3):431–479, 2011

  39. [47]

    C.-H. Zhang. Nearly unbiased variable selection under minimax concave penalty. The Annals of Statistics, 38(2):894 – 942, 2010

  40. [48]

    Zou and T

    H. Zou and T. Hastie. Regularization and variable selection via the elastic net. Journal of the royal statistical society: series B (statistical methodology) , 67(2):301–320, 2005

  41. [49]

    Sapienza

    H. Zou and R. Li. One-step sparse estimates in nonconcave penalized likelihood models. The Annals of Statistics , 36(4):1509 – 1533, 2008. Department of Statistical Sciences, “Sapienza” University of Rome, P.le Aldo Moro, 5 - 00185, Rome, Italy Email address : alessandro.degre...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.