Pith. sign in

REVIEW 3 major objections 5 minor 57 references

A Koopman-based framework for forecasting the spatiotemporal evolution of chaotic dynamics with nonlinearities modeled as exogenous forcings

T0 review · 3 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 A Koopman-based method that treats nonlinearities as external forcings predicts spatiotemporal chaos for several Lyapunov timescales.

desk verdict A useful physics-informed hybrid for short-term forecasting of chaos, honestly labeled as data-driven only when the forcing dictionary is right; worth refereeing with a requested robustness analysis. read the letter →

arxiv 1909.00076 v1 pith:HG4N4CB3 submitted 2019-08-30 physics.flu-dyn physics.comp-ph

Mohammad Amin Khodkar , Pedram Hassanzadeh , Athanasios Antoulas This is my paper · ORCID
classification physics.flu-dynphysics.comp-ph
keywords Koopmanoperatordata-drivenforecastingchaoticdynamicsKuramoto-SivashinskyequationLorenz-96systemdelayembeddingdynamicmodedecompositionwithcontrolexogenousforcing
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 proposes a data-driven method, called M2, for forecasting high-dimensional chaotic dynamics by approximating the Koopman operator on delay-embedded, vector-valued observables and treating the system's nonlinearity as an exogenous forcing term. The state update is linear in the augmented state $u_{m+1}=A u_m + B f_m$, where $f_m$ is constructed from physical knowledge of the governing equations, and $A$ and $B$ are learned jointly from data using dynamic mode decomposition with control. On the Kuramoto-Sivashinsky equation, Lorenz-96 system, and a 2D lid-driven cavity flow, the method yields accurate spatiotemporal forecasts for several Lyapunov timescales, with prediction horizons above $8/\Lambda_{\max}$ for moderately chaotic K-S systems and average relative errors below 7 percent. The performance modestly exceeds a reservoir computing baseline. The paper shows that a few growing linear modes are stabilized by the nonlinear forcing term, and that delay-embedding is essential, as setting the embedding dimension to one collapses the horizon.

What carries the argument

The machinery is a delay-embedded state space: the observable is a window of $q$ consecutive snapshots of vector-valued measurements, arranged in a Hankel matrix, and the approximated Koopman operator acts on a rank-$r$ subspace of that matrix. The forcing extension appends a block of nonlinear functions $f$ (e.g., squared velocity components for K-S, neighboring-point products for Lorenz-96, Reynolds stresses for the cavity flow) to form an augmented data matrix, and DMDc learns the reduced linear map $A$ and forcing map $B$ by solving a least-squares problem. The key operational rule from the paper is that the optimal delay $q$ satisfies $q\tau = O(\tau_d)$, where $\tau_d$ is the decorrelation time, and that the optimal truncation rank $r$ tracks the singular-value hard threshold, with the forcing rank $p$ set proportional to the ratio of forcing-vector length to state-vector length.

What would settle it

Recompute the case labeled $J=0$ in Fig. 4(a) for Lorenz-96 with $F=8$: use the same training data and method parameters but construct the forcing vector from only $X_j^2$, with no neighboring-point products. The paper reports a short horizon; if the prediction horizon instead matches the full-forcing case, the claim that cross-term forcings are essential would be contradicted.

Watch

Extended reading notes

Core claim

The central claim is that a finite-dimensional linear approximation of the Koopman operator can be made to predict chaotic spatiotemporal dynamics accurately for several Lyapunov timescales, provided the nonlinearity is represented as an exogenous forcing term of known functional form. Using delay-embedded vector observables and the DMDc algorithm to learn the pair $(A,B)$, the model $u_{m+1}=A u_m + B f_m$ tracks the true trajectory across systems ranging from Lorenz-63 (more than 10 Lyapunov times with 2.10 percent average error) to the K-S equation with domain length up to 1600 and a 2D cavity flow at Reynolds numbers 20000 and 30000. The paper emphasizes that the linear operator $A$ of M2 generically possesses a few eigenvalues outside the unit circle, and that the forcing term $Bf$ suppresses the unbounded growth of these modes, analogous to the stabilizing role of nonlinear energy transfer in turbulence. Without the forcing term, the Hankel-DMD baseline (M1) produces predictions that decay to zero within about a Lyapunov time.

Load-bearing premise

The method assumes the user can specify the functional form of the nonlinear forcing term; for Lorenz-96, including only $X_j^2$ yields short prediction horizons, and the paper states that the highest-order term needed can be determined by the investigator's speculation or intuition.

Editorial extensions

If this is right

  • For K-S systems with domain length $L \le 200$, predictions remain accurate for more than $8/\Lambda_{\max}$ with average error below 7 percent, which modestly exceeds the roughly $6/\Lambda_{\max}$ horizon of a reservoir computing baseline.
  • For the Lorenz-96 system with $F=8$, the prediction horizon is $8.16/\Lambda_{\max}$ with 6.82 percent average error, and for $F=4$ the predictions never diverge within the test window, indicating the method can sometimes recover quasiperiodic dynamics exactly.
  • For the 2D lid-driven cavity flow, the method predicts the vorticity field accurately for 5.2 advective timescales at $Re=20000$ and 2.7 advective timescales at $Re=30000$, corresponding to hundreds of DNS time steps.
  • Because the forcing terms are updated from the predicted state at each step, the trained model can be iterated forward in a closed loop without access to the true state.

Reading between the lines

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

  • We infer that the method's reliance on user-specified forcing terms could be relaxed by combining its PCC-based neighbor selection with sparse regression over candidate monomials; the paper leaves the order of nonlinearity to 'investigator's speculation,' which is the main human input.
  • If the forced linear representation is as general as the K-S and Lorenz-96 results suggest, then the practical bottleneck for data-driven forecasting of high-dimensional chaos may be identifying the correct forcing subspace rather than fitting a nonlinear evolution operator; this hypothesis is testable by comparing M2 with a deep learning model on the same benchmarks.
  • The authors' speculation about compressing 3D turbulence with autoencoders before applying M2 is a natural extension; a concrete next step would be to train an autoencoder on isotropic turbulence and check whether the latent-space forced linear model retains the predictive horizon seen in the 2D cavity flow.
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

3 major / 5 minor

Summary. The paper introduces a data-driven spatiotemporal forecasting framework, M2, based on a finite-dimensional Koopman approximation. Vector-valued delay-embedded observables form the state, while nonlinear terms are modeled as exogenous forcing functions f in the linear state-space model u_{m+1} = A u_m + B f_m, with A and B estimated using dynamic mode decomposition with control (DMDc). The forcing dictionary is chosen using physical knowledge: u^2 for the Kuramoto-Sivashinsky equation, monomials X_j X_{j+k} for Lorenz-96, and Reynolds stresses for the lid-driven cavity flow. The method is evaluated on Lorenz-63, K-S equations with domain lengths L = 22 to 1600, Lorenz-96 with F = 4, 8, and 16, and 2D lid-driven cavity flow at Re = 20000 and 30000. Forecast horizons are reported in Lyapunov timescales for the chaotic systems and advective timescales for the cavity flow, with the central claim that M2 provides accurate forecasts for several Lyapunov timescales and modestly outperforms the reservoir computing approach of Pathak et al.

Significance. If the reported forecast horizons are representative, M2 is a useful contribution to short-term spatiotemporal forecasting of high-dimensional chaotic systems. The paper has clear strengths: evaluation on twenty independent test sets, direct comparison with a recent reservoir computing benchmark, physically interpretable forcing construction, and an eigenvalue analysis that gives a mechanistic picture of unstable modes balanced by nonlinear terms. The principal limitation is that the method's skill depends on a correctly specified forcing dictionary that is supplied by the user rather than learned from data, which makes the 'data-driven' framing incomplete. That limitation is acknowledged in the text but not quantified or stress-tested, and it materially affects the scope of the central claim.

major comments (3)
  1. [Sec. II.B, Eq. (12), and Fig. 4(a)] The predictive skill of M2 depends critically on the user-supplied forcing dictionary, which is not learned from data. The paper states after Eq. (12) that the highest-order term required for constructing the forcing vector 'can be determined by the investigator's speculation or intuition,' and Fig. 4(a) shows that for Lorenz-96 with F = 16, omitting the J >= 2 cross terms (J = 0) drastically reduces the prediction horizon. For the K-S equation, the choice f_i = u_i^2 is motivated by the identity u u_x = (u^2)_x / 2, which is equation knowledge rather than a data-driven discovery. As written, the headline claim of a 'data-driven' forecasting framework is therefore overstated: the reported horizons are conditional on an external physical prior about the functional form of the nonlinearity. I recommend either adding a systematic study of dictionary misspecification (for example, random subsets of monomials or missing cross terms) or explicitly reframing the contribution as a physics-assisted Koopman forecasting method.
  2. [Table I and Sec. III.B] The comparison with the reservoir computing results of Pathak et al. and the statement that M2 'modestly outperforms' that approach are based on point estimates without uncertainty quantification. Although errors are averaged over 20 testing sets, no standard deviations, confidence intervals, or per-realization spreads are reported for t_l or E_ave. Because the best parameters (r, p, q) are selected by a validation search, the reported values are selected maxima and are susceptible to selection bias; the margin over Pathak et al. (roughly 8/Lambda_max versus 6/Lambda_max) may lie within the spread across realizations. Reporting the distribution of t_l and E_ave over the 20 test sets, and ideally over bootstrapped training sets, is necessary to support the comparative claim.
  3. [Figs. 12-14 and Sec. III.D] For the lid-driven cavity flow the averaged errors are high: E_ave = 16.6% at Re = 20000 and 16.8% at Re = 30000, with prediction horizons of 5.2 and 2.7 advective timescales. The abstract and conclusions describe the performance as 'accurate' and 'similar' to the chaotic test cases, but a 16-17% mean relative error is not in the same range as the 3-11% errors reported for the K-S and Lorenz systems, and the cavity results are not compared with a baseline method. The claim should either be qualified with a discussion of what 'accurate' means for this application or supported by an error-versus-time curve analogous to Fig. 2(b) and a comparison with M1 or another baseline.
minor comments (5)
  1. [Abstract and Sec. I] There are multiple grammatical and typographical errors, including 'introduce a data-driven method and shows' in the abstract and 'an reservoir computing' in the introduction; a careful proofread is needed.
  2. [Eq. (10) and surrounding text] The dimensions and roles of the SVD factors in the DMDc formulas would benefit from a more explicit statement; in particular, the partition of U-tilde into U-tilde_1 and U-tilde_2 is only clear after rereading the derivation.
  3. [Fig. 13 caption] The caption says 'Similar to Fig. 13' but the intended cross-reference is clearly Fig. 12.
  4. [Sec. III.D, last paragraph] The sentence claiming that q*tau is approximately 0.2*tau_d for the cavity flow appears inconsistent with the stated values tau approximately 0.025*tau_d and q = 3, which give q*tau approximately 0.075*tau_d; please check the reported value.
  5. [Throughout] There are numerous typographical errors (for example, 'choticity', 'whcih', 'unergo', 'compoenents', 'beahaviour', 'fucntion', 'fro cing', 'sudies') that should be corrected before publication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: M2 fits A and B from training data and evaluates on independent testing sets; the physics-based forcing dictionary is an external input rather than a fitted prediction.

full rationale

The core derivation chain is not circular. Equations (7)-(10) define M2 by fitting the linear maps A and B with DMDc from training data, and all reported forecast horizons are evaluated on testing sets that are explicitly separate from training and validation data (Sec. II.A, Fig. 2, Tables I-II). The forcing vectors are not learned from the target predictions; they are supplied from physical knowledge or from the PCC heuristic in Eq. (13), and the paper is explicit that the highest-order term is chosen by investigator intuition (Sec. II.B). This makes the method conditional on an external prior about the nonlinearity, but it does not make the prediction equivalent to its inputs by construction. The failure of the J=0 Lorenz-96 forcing (Fig. 4a) is a sensitivity result, not a circular reduction. Hyperparameters (r,p,q) are selected on independent validation sets rather than on the testing sets used to report skill. Self-citations, such as the Ionita-Antoulas matrix-pencil correspondence [40], are contextual and not load-bearing for the central forecast claim, which rests on external methods (Hankel-DMD [30], DMDc [45]) and an external benchmark (Pathak et al. [18]). The comparison with Pathak et al. uses matched training data and independent test data, so the reported outperformance is a genuine out-of-sample result rather than a renamed fit.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The method's predictive skill rests on the user providing the right nonlinear dictionary; the paper tunes ranks and delays on validation data. No new physical entities are introduced.

free parameters (6)
  • delay-embedding dimension q = 10 to 50 (e.g., 20 for K-S L=22, 40 for Lorenz-96 F=16, 3 for cavity)
    Chosen by comprehensive search on validation sets (Appendix A).
  • reduced rank r = 400 to 16000 (e.g., 400 for K-S L=22, 1200 for Lorenz-96 F=16, 400/1200 for cavity)
    Chosen by validation search (Appendix A).
  • augmented rank p = 400 to 16000, often p = α r
    Chosen by validation search; α is the ratio of forcing length to state length (Appendix A).
  • number of neighboring points J in Lorenz-96 forcing = 2 (for F=8 and 16)
    Selected based on abrupt jump in tl vs J (Fig. 4a) and PCC threshold 0.1 (Sec. II.B).
  • PCC threshold = 0.1
    Chosen by hand to identify neighboring points for forcing construction (Sec. II.B).
  • forcing dictionary = u^2 for K-S; X_j^2 and X_j X_{j+k} for k=1..J for Lorenz-96; u'^2, u'v', v'^2 for cavity
    Chosen based on physical knowledge of each system; not derived from data (Sec. II.A, II.B, III.D).
assumptions (5)
  • standard math Koopman operator can be approximated by a finite-dimensional linear operator on delay-embedded observables.
    Invoked in Sec. II.A, citing Arbabi and Mezić 2017 and Takens embedding theorem.
  • domain assumption The linear-plus-forcing model u_{m+1}=A u_m + B f_m can accurately represent the chaotic dynamics on the training attractor.
    This is the central modeling assumption; no proof is given that such a low-rank linear model with the chosen f suffices.
  • ad hoc to paper The chosen forcing functions (u^2, monomials, Reynolds stresses) capture the essential nonlinearity of each system.
    Chosen by physics intuition; for Lorenz-96, J=0 (only X_j^2) fails, showing sensitivity to this choice (Sec. II.B).
  • domain assumption Training and testing data are drawn from the same stationary attractor and are independent.
    Standard for data-driven forecasting; the paper uses separate training and testing sets (Sec. II.A).
  • domain assumption For the cavity flow, removing the long-time mean and using anomalous velocities as state is sufficient.
    Stated in Sec. III.D; assumes statistics are stationary.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Koopman-based framework for forecasting the spatiotemporal evolution of chaotic dynamics with nonlinearities modeled as exogenous forcings." pith.science (2026). https://pith.science/paper/HG4N4CB3

@misc{pith2026190900076,
  author       = {Pith},
  title        = {Pith review of: A Koopman-based framework for forecasting the spatiotemporal evolution of chaotic dynamics with nonlinearities modeled as exogenous forcings},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HG4N4CB3}},
  note         = {Machine review of arXiv:1909.00076}
}
read the original abstract

We introduce a data-driven method and shows its skills for spatiotemporal prediction of high-dimensional chaotic dynamics and turbulence. The method is based on a finite-dimensional approximation of the Koopman operator where the observables are vector-valued and delay-embedded, and the nonlinearities are treated as external forcings. The predictive capabilities of the method are demonstrated for well-known prototypes of chaos such as the Kuramoto-Sivashinsky equation and Lorenz-96 system, for which the data-driven predictions are accurate for several Lyapunov timescales. Similar performance is seen for two-dimensional lid-driven cavity flows at high Reynolds numbers.

Figures

Figures reproduced from arXiv: 1909.00076 by the authors.

Figure 1
Figure 1. FIG. 1. Schematic of data-driven method M2: (a) Training on vector-valued and time-delay [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. (a) Predictions given by M1 (dashed magenta) and M2 (dashed red) for the time series of [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Eigenvalues [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: FIG. 4. (a) Variation of prediction horizon [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Predictions of the present method M2 for the temporal evolution of the variables in a [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Eigenvalues [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7. Spatiotemporal evolution of an unforced K-S system with the domain length [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Spatiotemporal evolution of a Lorenz-96 system with the external forcing [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]
Figure 9
Figure 9. Figure 9: displays the eigenvalues of the Lorenz-96 system with F = 8 given by M1 (left) and M2 (right). As demonstrated by this figure, not surprisingly, all modes of AHDMD are decaying again, leading to the predictions that approach zero fairly rapidly, and lose the true traje…
Figure 10
Figure 10. Figure 10: The constant-density fluid is confined by a square box whose walls are stationary, [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]
Figure 10
Figure 10. Figure 10: FIG. 10. Schematic of a 2D lid-driven cavity flow. Here, [PITH_FULL_IMAGE:figures/full_fig_p021_10.png]
Figure 11
Figure 11. Figure 11: FIG. 11. The Power Spectral Densities (PSDs) of the time series of the PC1 calculated for the [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]
Figure 12
Figure 12. Figure 12: FIG. 12. Snapshots of vorticity field for a lid-driven cavity flow at Re = 20000 at six distinct [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]
Figure 13
Figure 13. Figure 13: FIG. 13. Similar to Fig. 13, but for a flow at Re = 30000. The optimal method parameters are [PITH_FULL_IMAGE:figures/full_fig_p024_13.png]
Figure 14
Figure 14. Figure 14: FIG. 14. Timeseries of the PC1 of horizontal velocity, calculated from DNS (solid blue) and [PITH_FULL_IMAGE:figures/full_fig_p025_14.png]
Figure 15
Figure 15. Figure 15: FIG. 15. Eigenvalues [PITH_FULL_IMAGE:figures/full_fig_p026_15.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 53 canonical work pages

  1. [1]

    Deterministic nonperiodic flow,

    E. N. Lorenz, “Deterministic nonperiodic flow,” J. Atoms. Sci. 20, 130–141 (1963)

    work page 1963

  2. [2]

    ˆU ∈ Rnq×r, ˆS ∈ Rr×r and ˆV∈ R(N−q)×r, and [X F]T = ˜U˜S˜V∗ with the truncation value selected as p, so that ˜U∈ R(n+n′)q×p, ˜S∈ Rp×p and ˜V∈ R(N−q)×p

    (10) Here, Y = ˆUˆSˆV∗, while the truncation value is taken as r, i.e. ˆU ∈ Rnq×r, ˆS ∈ Rr×r and ˆV∈ R(N−q)×r, and [X F]T = ˜U˜S˜V∗ with the truncation value selected as p, so that ˜U∈ R(n+n′)q×p, ˜S∈ Rp×p and ˜V∈ R(N−q)×p. ˜U1 and ˜U2 are made up of the first nq and the remaining n′q rows of ˜U, respectively. Note that A and B are calculated in a reduced-...

    work page

  3. [3]

    G. E. P. Box, Jenkins G. M., G. C. Reinsel, and G. M. Ljung, Time series analysis: forecasting and control (John Wiley & Sons, 2015)

    work page 2015

  4. [4]

    Long-term research challenges in wind energy - A research agenda by the European academy of wind energy,

    G. Van Kuik et al., “Long-term research challenges in wind energy - A research agenda by the European academy of wind energy,” Wind Energ. Sci. 1, 1–39 (2016)

    work page 2016

  5. [5]

    Duriez, S

    T. Duriez, S. L. Brunton, and B. R. Noack, Machine Learning Control-Taming Nonlinear Dynamics and Turbulence (Springer, 2017)

    work page 2017

  6. [6]

    Challenges in climate science and contemporary applied mathematics,

    A. J. Majda, “Challenges in climate science and contemporary applied mathematics,” Com- mun. Pure Appl. Math. 65, 920–948 (2012)

    work page 2012

  7. [7]

    The quiet revolution of numerical weather prediction,

    P. Bauer, A. Thorpe, and G. Brunet, “The quiet revolution of numerical weather prediction,” Nature 525, 47 (2015)

    work page 2015

  8. [8]

    Model error, information barriers, state estimation and prediction in complex multiscale systems,

    A. Majda and N. Chen, “Model error, information barriers, state estimation and prediction in complex multiscale systems,” Entropy 20, 644 (2018)

    work page 2018

Show all 57 references
  1. [9]

    A variational approach to probing extreme events in turbu- lent dynamical systems,

    M. Farazmand and T. P. Sapsis, “A variational approach to probing extreme events in turbu- lent dynamical systems,” Sci. Adv. 3 (2017)

    2017

  2. [10]

    The interpretation of short climate records, with comments on the North Atlantic and Southern Oscillations,

    C. Wunsch, “The interpretation of short climate records, with comments on the North Atlantic and Southern Oscillations,” Bull. Am. Meteorol. Soc. 80, 245–256 (1999)

    1999

  3. [11]

    Van den Dool, Empirical methods in short-term climate prediction (Oxford University Press, 2007)

    H. Van den Dool, Empirical methods in short-term climate prediction (Oxford University Press, 2007)

    2007

  4. [12]

    The skill of atmospheric linear inverse models in hindcasting the Madden–Julian Oscillation,

    N. R. Cavanaugh, T. Allen, A. Subramanian, B. Mapes, H. Seo, and A. J. Miller, “The skill of atmospheric linear inverse models in hindcasting the Madden–Julian Oscillation,” Clim. Dyn. 44, 897–906 (2015)

    2015

  5. [13]

    Data-driven spectral decomposition and forecasting of ergodic dynamical sys- tems,

    D. Giannakis, “Data-driven spectral decomposition and forecasting of ergodic dynamical sys- tems,” Appl. Comput. Harmon. Anal. (2017)

    2017

  6. [14]

    Data-driven prediction strategies for low-frequency patterns of North Pacific climate variability,

    D. Comeau, Z. Zhao, D. Giannakis, and A. J. Majda, “Data-driven prediction strategies for low-frequency patterns of North Pacific climate variability,” Clima. Dyn. 48, 1855–1872 (2017). 29

    2017

  7. [15]

    Data-driven reduced modelling of turbulent Rayleigh- B´ enard convection using dmd-enhanced fluctuation-dissipation theorem,

    M. A. Khodkar and P. Hassanzadeh, “Data-driven reduced modelling of turbulent Rayleigh- B´ enard convection using dmd-enhanced fluctuation-dissipation theorem,” J. Fluid Mech.852 (2018), 10.1017/jfm.2018.586

    2018 doi

  8. [16]

    An ensemble quadratic echo state network for non-linear spatio-temporal forecasting,

    P. L. McDermott and C. K. Wikle, “An ensemble quadratic echo state network for non-linear spatio-temporal forecasting,” Stat 6, 315–330 (2017)

    2017

  9. [17]

    Long-term forecasting using tensor-train RNNs,

    R. Yu, S. Zheng, A. Anandkumar, and Y. Yue, “Long-term forecasting using tensor-train RNNs,” arXiv:1711.00073 (2017)

    2017 arXiv

  10. [18]

    Data-driven fore- casting of high-dimensional chaotic systems with long short-term memory networks,

    P. R. Vlachas, W. Byeon, Z. Y. Wan, T. Sapsis, and P. Koumoutsakos, “Data-driven fore- casting of high-dimensional chaotic systems with long short-term memory networks,” Proc. R. Soc. A 474 (2018)

    2018

  11. [19]

    Model-free prediction of large spatiotem- porally chaotic systems from data: A reservoir computing approach,

    U. Pathak, B. Hunt, M. Girvan, Z. Lu, and E. Ott, “Model-free prediction of large spatiotem- porally chaotic systems from data: A reservoir computing approach,” Phys. Rev. Lett. 120 (2018)

    2018

  12. [20]

    Physics-informed neural networks: A deep learning framework for learning forward and inverse problems involving nonlinear partial differential equations,

    M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Physics-informed neural networks: A deep learning framework for learning forward and inverse problems involving nonlinear partial differential equations,” J. Comp. Phys. 378, 686–707 (2019)

    2019

  13. [21]

    Compressed convolutional LSTM: An efficient deep learning framework to model high fidelity 3D turbulence,

    A. T. Mohan, D. Daniel, M. Chertkov, and D. Livescu, “Compressed convolutional LSTM: An efficient deep learning framework to model high fidelity 3D turbulence,” arXiv:1903.00033 (2019)

    2019 arXiv

  14. [22]

    Deep state networks with uncertainty quantification for spatio-temporal forecasting,

    P. L. McDermott and C. K. Wikle, “Deep state networks with uncertainty quantification for spatio-temporal forecasting,” Environmetrics 30, e2553 (2019)

    2019

  15. [23]

    Data-driven prediction of a multi-scale lorenz96 chaotic system using a hierarchy of deep learning methods: Reservoir computing, ANN, and RNN-LSTM,

    A. Chattopadhyay, P. Hassanzadeh, K. Palem, and D. Subramanian, “Data-driven prediction of a multi-scale lorenz96 chaotic system using a hierarchy of deep learning methods: Reservoir computing, ANN, and RNN-LSTM,” arXiv:1906.08829 (2019)

    2019 arXiv

  16. [24]

    Hamiltonian systems and transformation in Hilbert space,

    B. O. Koopman, “Hamiltonian systems and transformation in Hilbert space,” Proc. Natl. Acad. Sci. 17 (1931)

    1931

  17. [25]

    Spectral properties of dynamical systems, model reduction and decompositions,

    I. Mezi´ c, “Spectral properties of dynamical systems, model reduction and decompositions,” Nonlin. Dyn. 41, 309–325 (2005)

    2005

  18. [26]

    Analysis of fluid flows via spectral properties of the Koopman operator,

    I. Mezi´ c, “Analysis of fluid flows via spectral properties of the Koopman operator,” Annu. Rev. Fluid Mech. 45 (2013). 30

    2013

  19. [27]

    Dynamic mode decomposition of numerical and experimental data,

    P. J. Schmid, “Dynamic mode decomposition of numerical and experimental data,” J. Fluid Mech. 656, 5–28 (2010)

    2010

  20. [28]

    Spectral analysis of nonlinear flows,

    C. W. Rowley, I. Mezi´ c, S. Bagheri, P. Schlatter, and D. S. Henningson, “Spectral analysis of nonlinear flows,” J. Fluid Mech. 641, 115–127 (2009)

    2009

  21. [29]

    On dynamic mode decomposition: Theory and applications,

    J. H. Tu, C. W. Rowley, Luchtenburg D. M., S. L. Brunton, and J. N. Kutz, “On dynamic mode decomposition: Theory and applications,” J. Comp. Dyn. 1, 391–421 (2014)

    2014

  22. [30]

    A data–driven approximation of the Koopman operator: Extending dynamic mode decomposition,

    M. O. Williams, I. G. Kevrekidis, and C. W. Rowley, “A data–driven approximation of the Koopman operator: Extending dynamic mode decomposition,” J. Nonlin. Sci. 25 (2015)

    2015

  23. [31]

    Ergodic theory, dynamic mode decomposition, and computation of spectral properties of the Koopman operator,

    H. Arbabi and I. Mezi´ c, “Ergodic theory, dynamic mode decomposition, and computation of spectral properties of the Koopman operator,” SIAM J. Appl. Dyn. Syst. 16, 2096–2126 (2017)

    2017

  24. [32]

    Study of dynamics in post-transient flows using Koopman mode decomposition,

    H. Arbabi and I. Mezi´ c, “Study of dynamics in post-transient flows using Koopman mode decomposition,” Phys. Rev. Fluids 2, 124402 (2017)

    2017

  25. [33]

    On convergence of extended dynamic mode decomposition to the Koopman operator,

    M. Korda and I. Mezi´ c, “On convergence of extended dynamic mode decomposition to the Koopman operator,” J. Nonlin. Sci. 28, 687–710 (2018)

    2018

  26. [34]

    Model reduction for flow analysis and control,

    C. W. Rowley and S. T. M. Dawson, “Model reduction for flow analysis and control,” Annu. Rev. Fluid Mech. 49, 387–417 (2017)

    2017

  27. [35]

    Detecting strange attractors in turbulence,

    F. Takens, “Detecting strange attractors in turbulence,” Lect. Notes Math. 898, 366–381 (1981)

    1981

  28. [36]

    Linear predictors for nonlinear dynamical systems: Koopman oper- ator meets model predictive control,

    M. Korda and I. Mezi´ c, “Linear predictors for nonlinear dynamical systems: Koopman oper- ator meets model predictive control,” Automatica 93, 149–160 (2018)

    2018

  29. [37]

    A data-driven Koopman model predictive control for nonlinear flows,

    H. Arbabi, M. Korda, and I. Mezi´ c, “A data-driven Koopman model predictive control for nonlinear flows,” arXiv:1804.05291 (2018)

    2018 arXiv

  30. [38]

    Nonlinear laplacian spectral analysis for time series with intermittency and low-frequency variability,

    D. Giannakis and A. J. Majda, “Nonlinear laplacian spectral analysis for time series with intermittency and low-frequency variability,” Proc. Natl Acad. Sci. 113, 3932–3937 (2012)

    2012

  31. [39]

    Chaos as an intermittently forced linear system,

    S. L. Brunton, B. W. Brunton, J. L. Proctor, E. Kaiser, and J. N. Kutz, “Chaos as an intermittently forced linear system,” Nat. Commun. 8, 19 (2017)

    2017

  32. [40]

    Discovering governing equations from data by sparse identification of nonlinear dynamical systems,

    S. L. Brunton, J. L. Proctor, E. Kaiser, and J. N. Kutz, “Discovering governing equations from data by sparse identification of nonlinear dynamical systems,” Proc. Natl Acad. Sci.113, 3932–3937 (2016). 31

    2016

  33. [41]

    Matrix pencils in time and frequency domain system identification,

    A. C. Ionita and A. C. Antoulas, “Matrix pencils in time and frequency domain system identification,” in Developments in Control Theory: Towards Glocal Control , Vol. 76, edited by L. Qiu, J. Chen, T. Iwasaki, and H. Fujioka (IET Control Engineering Series, 2012) pp. 79–88

    2012

  34. [42]

    Model reduction of bilinear systems in the Loewner framework,

    A. C. Antoulas, I. V. Gosea, and A. C. Ionita, “Model reduction of bilinear systems in the Loewner framework,” SIAM J. Sci. Comput. 38, B889–B916 (2016)

    2016

  35. [43]

    On the Loewner framework for model reduction of Burgers’ equation,

    A. C. Antoulas, I. V. Gosea, and M. Heinkenschloss, “On the Loewner framework for model reduction of Burgers’ equation,” inActive Flow and Combustion Control 2018 (Springer, 2019) pp. 255–270

    2018

  36. [44]

    Gugercin, C

    S. Gugercin, C. A. Beattie, and A. C. Antoulas, Data-driven and interpolatory model reduction (Society for Industrial and Applied Mathematics, 2019)

    2019

  37. [45]

    Clustering of series via dynamic mode decomposition and the matrix pencil method,

    L. Pogorelyuk and C. W. Rowley, “Clustering of series via dynamic mode decomposition and the matrix pencil method,” arXiv:1802.09878 (2018)

    2018 arXiv

  38. [46]

    Dynamic mode decomposition with control,

    J. L Proctor, S. L. Brunton, and J. N. Kutz, “Dynamic mode decomposition with control,” SIAM J. Appl. Dynam. Syst. 15, 142–161 (2016)

    2016

  39. [47]

    Determining Lyapunov exponents from a time series,

    A. Wolf, J. B. Swift, H. L. Swinney, and J. A. Vastano, “Determining Lyapunov exponents from a time series,” Physica D 16, 285–317 (1985)

    1985

  40. [48]

    Statistically accurate low-order models for uncertainty quantifi- cation in turbulent dynamical systems,

    T. Sapsis and A. J. Majda, “Statistically accurate low-order models for uncertainty quantifi- cation in turbulent dynamical systems,” Proc. Natl. Acad. Sci. 110 (2013)

    2013

  41. [49]

    Introduction to turbulent dynamical systems for complex systems,

    A. J. Majda, “Introduction to turbulent dynamical systems for complex systems,” in Frontiers in Applied Dynamical Systems: Reviews and Tutorials (Springer, 2016)

    2016

  42. [50]

    Low-dimensional reduced-order models for statistical response and uncertainty quantification: Two-layer baroclinic turbulence,

    D. Qi and A. J. Majda, “Low-dimensional reduced-order models for statistical response and uncertainty quantification: Two-layer baroclinic turbulence,” J. Atoms. Sci. 73, 4609–4639 (2016)

    2016

  43. [51]

    J. L. Kaplan and J. A. Yorke, Functional Differential Equations and the Approximation of Fixed Points, Lecture Notes in Mathem, Vol. 730 (Springer, 1978) pp. 204–227

    1978

  44. [52]

    Predictability - a problem partly solved,

    E. N. Lorenz, “Predictability - a problem partly solved,” in Predictability of Weather and Climate, edited by T. Palmer and R. Hagedorn (Cambridge University Press, 2006) pp. 40– 58

    2006

  45. [53]

    High-Re solutions for incompressible flow using the Navier-Stokes equations and a multigrid method,

    U. Ghia, N. K. Ghia, and C. T. Shin, “High-Re solutions for incompressible flow using the Navier-Stokes equations and a multigrid method,” J. Comput. Phys. 48, 387–411 (1982). 32

    1982

  46. [54]

    Driven cavity flows by efficient numerical techniques,

    H. B. Schreiber and H. B. Keller, “Driven cavity flows by efficient numerical techniques,” J. Comput. Phys. 49, 310–333 (1983)

    1983

  47. [55]

    A novel fully implicit finite volume method applied to the lid- driven cavity problempart I: High Reynolds number fow calculations,

    M. Sahin and R. G. Owens, “A novel fully implicit finite volume method applied to the lid- driven cavity problempart I: High Reynolds number fow calculations,” J. Numer. Meth. Fluids 42, 57–77 (2003)

    2003

  48. [56]

    Linearly recurrent autoencoder networks for learning dynamics,

    S. E. Otto and C. W Rowley, “Linearly recurrent autoencoder networks for learning dynamics,” SIAM J. Sci. Comput. 18, 558–593 (2019)

    2019

  49. [57]

    Singular Value Decomposition (svd) and Principal Component Analysis (PCA),

    S. L. Brunton and J. N. Kutz, “Singular Value Decomposition (svd) and Principal Component Analysis (PCA),” in Data-driven Science and Engineering: Machine Learning, Dynamical Systems, and Control (Cambridge University Press, 2019) Chap. 1. 33

    2019

Pith tools

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