Pith. sign in

REVIEW 3 major objections 3 minor 89 references

Data-driven reconstruction of dynamical systems using Takens' Theorem, manifold learning, and universal function approximators

T0 review · 3 major / 3 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read The paper claims that integrating Takens delay embeddings, diffusion maps, and neural networks yields a general pipeline—TAR—that reconstructs a dynamical system's full state from one or a few low-dimensional time series, and reconstructs…

desk verdict Extension of STAR to arbitrary dynamical systems with a new multi-temporal embedding; solid work whose abstract overstates the VOO claims. read the letter →

arxiv 2608.05477 v1 pith:C6BY7C44 submitted 2026-08-05 physics.comp-ph

Maximilian Topel , Andrew L. Ferguson This is my paper · ORCID
classification physics.comp-ph
keywords Takens'theoremdelayembeddingdiffusionmapsmanifoldlearningneuralnetworksstate-spacereconstructionmulti-temporaltimeseriesanalysis
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

TAkens Reconstruction (TAR) is a proposed answer to how to go from the theory of Takens' delay embedding to a working algorithm: organize one or more time series into delay vectors, learn the intrinsic manifold with diffusion maps, and train neural networks to map that manifold to the full system state. The paper's claim is that this pipeline reconstructs the high-dimensional state of a dynamical system from low-dimensional observations—prey population from predator counts, protein coordinates from an end-to-end distance—and, when full-dimensional training data do not exist, reconstructs predictive distributions over observables, as demonstrated on VOO share-price returns. The authors also claim that delay-vector structure matters: total time span $p\tau$ rather than a single optimal $\tau$ controls accuracy, and multi-temporal embeddings built from several delays improve reconstruction for systems with multiple time scales. If these claims hold, experimentalists and analysts who can record only one or a few time series gain a general, nonparametric route to the hidden state and dynamics of the system.

What carries the argument

The central object is Takens' delay vector $y(t)=[O(t),O(t-\tau),\ldots,O(t-(p-1)\tau)]$, treated as a short trajectory snippet that carries enough history to pin down the system's current state. Diffusion maps perform spectral decomposition of a diffusion operator to turn these delay vectors into coordinates on the intrinsic manifold $M'$, and a second application does the same for full-dimensional training data, yielding $M$. Fully connected feed-forward neural networks then approximate the smooth bijection between the two coordinate systems and the inverse of the diffusion-map reduction, which is the lifting back to high-dimensional states. Multi-temporal embeddings concatenate delay vectors built at several different $\tau$ values so that one observable can carry information from several time scales at once.

What would settle it

Train TAR on a Lorenz system sampled from one lobe only, then feed a trajectory that enters the other lobe; if reconstruction collapses in the unseen lobe, the assumption that test delay vectors lie on the training manifold is falsified.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the abstract guarantee of Takens' theorem—that delay vectors of a generic observable are diffeomorphic to the full state—can be turned into a concrete, data-driven reconstruction path. Delay vectors $y(t)=[O(t),O(t-\tau),\ldots,O(t-(p-1)\tau)]$ with $p\ge 2k+1$ are embedded by diffusion maps into an image $M'$ of the true intrinsic manifold $M$; where full-dimensional training trajectories are available, a second diffusion map learns $M$ and two feed-forward neural networks approximate the diffeomorphism $M'\to M$ and the lifting $M\to$ full state, so that a new low-dimensional time series can be fed through the path (A)→(B)→(C)→(E)→(F) to predict the full state. Without full-dimensional training data, the learned $M'$ alone supports backmapping to distributions, as in the VOO case where a TAR model trained on the first half of January 2024 reproduces the test-half return distribution with total-variation distances as low as 0.021 at the 1 min scale and 0.067 at 1 h, beating the parametric NIG, Gaussian, and GARCH(1,1)-t benchmarks at the two longer time scales.

Load-bearing premise

The load-bearing premise is that the delay vectors encountered after training lie on the same low-dimensional manifold as those used to train the diffusion map and neural networks, so that states visited only later—or new dynamical regimes—are still approximated by the learned geometry.

Editorial extensions

If this is right

  • For simple periodic systems, a delay vector must span roughly half a period or more; the common practice of picking $\tau$ at the first autocorrelation or mutual-information minimum did not coincide with the best reconstruction in Lotka-Volterra.
  • Multi-temporal embeddings let a single observable capture multiple time scales without delay vectors of astronomical dimension; in Villin, mt-5 improved reconstruction from 0.332 nm to 0.321 nm over the best single-delay model, though within 95% confidence intervals.
  • Adding a delay time whose single-temporal manifold is poorly learned can degrade a multi-temporal model, since mt-6 worsens relative to mt-5, so delay time selection, not just accumulation, matters.
  • On VOO, TAR recovers expected market regimes—mean reversion at seconds, random walk at minutes, trending at hours—and matches or beats NIG, Gaussian, and GARCH(1,1)-t at reproducing the test-half return distribution at $\tau=1$ min and $\tau=1$ h, with a roughly 40% gain at 1 h.
  • The open-source TAR code makes the pipeline usable on arbitrary time series, so the specific systems studied are demonstrations of a general workflow.

Reading between the lines

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

  • If TAR is right, a practical design rule for delay embeddings emerges that the paper only hints at: the total span $p\tau$ of a delay vector is the primary control, and conventional autocorrelation-based $\tau$ selection can be a poor proxy; this is testable on other periodic and chaotic systems.
  • The Villin finding that mt-5 beats single-delay models but within confidence intervals suggests the multi-temporal benefit is real but modest on one trajectory; a sharper test would apply multi-temporal TAR to several independent trajectories or to a system with well-separated time scales and look for consistent improvement.
  • For the VOO application, the absence of ground truth means the learned 'manifold' is asserted rather than verified; a natural stress test is to apply TAR to a synthetic market or agent-based model where the true latent state is known and compare the learned $M'$ to it.
  • Because Nyström out-of-sample projection lets new points be placed onto $M'$, TAR could serve as an anomaly or regime-shift detector: a point whose projection error is large is likely in a newly visited region of phase space, which could flag when the model needs retraining. The paper suggests this use but does not develop it.
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 / 3 minor

Summary. The paper introduces TAR, a computational pipeline that combines Takens' delay embedding, diffusion-map manifold learning, and feedforward neural networks to reconstruct full-dimensional dynamical states from low-dimensional time series. The method is demonstrated on three systems: a periodic Lotka-Volterra predator-prey model, the 125.6 microsecond molecular dynamics trajectory of the Villin protein, and tick-level returns of the Vanguard S&P 500 ETF. The authors report that TAR reconstructs the unobserved prey population with small error over a wide range of delay times, that multi-temporal delay embeddings modestly improve Villin reconstruction over single-temporal models, and that TAR can reproduce return distributions in the absence of full-dimensional market observations. An open-source software package and data are released.

Significance. If the central out-of-sample reconstruction claim holds, TAR is a potentially useful generic framework that operationalizes Takens' theorem with modern machine learning. The paper's strengths include the release of open-source code and data, a clean Lotka-Volterra parameter screen, Villin reconstructions that clearly beat the constant-structure baseline, and a VOO analysis that compares TAR against standard parametric benchmarks with held-out test partitions. However, the headline multi-temporal Villin improvement is within the paper's own 95% confidence intervals, and the out-of-sample validity of the learned manifold is asserted rather than demonstrated. The significance of the paper is therefore moderate: the framework is plausible and the demonstrations are suggestive, but the central generalization claim needs stronger quantitative support.

major comments (3)
  1. [§IV.B, §IV.C, §VI, §VII] The central predictive claim—that a new low-dimensional time series can be lifted to full states through the path Fig. 1(A)→(B)→(C)→(E)→(F)—requires test delay vectors to lie in the support of the training distribution on M', because the Nyström extension of an out-of-support point collapses toward the constant eigenvector and would feed uninformative coordinates to the subsequent ANNs. The paper provides no diagnostics for this. For Villin (§VI) the holdout is a terminal 20% contiguous block and only mean RMSD with 95% confidence intervals is reported; for VOO (§VII) Fig. 4(B)(iii) shows a substantial train/test distribution deviation at τ=1 h, yet the text asserts that the test dynamics are 'adequately represented within the training data' without quantitative support. I recommend adding a coverage or density diagnostic for projected test points—for example, the fraction of test embeddings in low-density regions of the training manifold, or reconstruction error stratified by training density—and reporting it for each experiment.
  2. [§VI, Fig. 3(A)] The conclusion that multi-temporal embeddings improve reconstruction beyond single-temporal TAR is not statistically supported. The best multi-temporal model mt-5 achieves 0.321 nm heavy-atom RMSD versus 0.332 nm for the best single-temporal model at τ=10 ns, and the paper itself states that this improvement lies within the 95% bootstrap confidence intervals. The monotonic mt-2 through mt-5 trend and the CCA complementarity analysis in Fig. 3(C) are suggestive, but they do not establish a significant improvement. The statement that the results demonstrate that multi-temporal delay embeddings can improve TAR performance should be rephrased, or supplemented with a formal comparison of the confidence intervals and an effect-size analysis.
  3. [§VII] The claim that TAR 'exposes' the mean-reverting, random-walk, and trend-following regimes in the VOO data is partially built in by construction: the delay times τ=2 s, 1 min, and 1 h are explicitly selected because they are expected to fall in those three regimes. The Hurst exponent analysis in Fig. 4(A) provides an independent check, but the paper should either present a scan over additional delay times or frame the VOO result as a consistency check with prior financial knowledge rather than as discovery of regimes. In addition, the assertion that the test-set dynamics are adequately represented within the training data should be supported by a quantitative measure, as noted in the first major comment.
minor comments (3)
  1. [§V] The delay-time list in the parameter screen reads '0.1.0.5' and should be '0.1, 0.5'; this is a typographical error that should be corrected.
  2. [§V] The text refers to the 'time span' of a delay vector as pτ, but a p-dimensional delay vector spans (p−1)τ; for example, p=3 and τ=0.001 gives a span of 0.002, not 0.003. The figure and text should use the corrected definition.
  3. [§IV.A and §VII] The manuscript states that TAR assumes uniformly sampled data, but the VOO analysis uses tick data coarsened to a 1-second grid by last-observation-carried-forward. This discrepancy should be acknowledged more explicitly in the main text rather than only in the analysis details.

Circularity Check

1 steps flagged · score 6.0 of 10

VOO return-distribution 'prediction' reduces to auto-encoding the target return that is already the first coordinate of each delay vector.

  1. fitted input called prediction [Section VII (Vanguard S&P 500 ETF), ANN backmapping description and Table I caption; cf. Eq. (1) in Section III.]
    "Y^j_t = [O^j_t, O^j_{t−τ}, ..., O^j_{t−(p−1)τ}] ... Finally, an ANN to learn a mapping from the attractor M′ to the returns distribution ... TAR is the ANN lift ψ→r trained on the train half and forward-evaluated through the Nyström extension on the test half."

    For VOO the observable time series O_t is the log return, so the delay vector's first coordinate is exactly the return r_t that the ANN ψ→r is trained to output. Each test delay vector therefore contains its own target as an input feature; the train/test split and Nyström out-of-sample projection do not alter this identity. The TV distances in Table I thus quantify how well the diffusion-map-plus-ANN auto-encodes the test returns, not how well TAR predicts an unobserved quantity. This is distinct from the Lotka-Volterra and Villin applications, where the reconstructed variables (prey population, all-atom coordinates) are absent from the delay-vector input.

full rationale

The core TAR mapping is genuinely predictive in the Lotka-Volterra and Villin applications: delay vectors are built from one observable (predator population or head-to-tail distance), and the target variables (prey population or all-atom coordinates) are not contained in the delay-vector input, with held-out test partitions used for evaluation. The VOO application, however, is circular in the specific sense that the ANN is trained to map the diffusion-map embedding of a delay vector back to the return series from which that delay vector was constructed; by Eq. (1), the return r_t is the first coordinate of the input delay vector Y_t. Consequently, the 'TAR lift reconstruction' of the test return distribution is a function of the very returns it claims to reconstruct, and the reported TV distances measure auto-encoding fidelity rather than predictive reconstruction. The self-citations to the authors' prior STAR work are routine and not load-bearing because the present paper supplies independent held-out validations. The choice of VOO delay times based on expected market regimes is not itself circular, since the regime structure is independently corroborated by Hurst exponents; it is a design choice rather than a derivation. Overall, the circularity is partial: one of three application domains reduces by construction, while the others stand as genuine reconstructions.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

TAR itself is a procedure, not an invented physical entity. The load-bearing choices are hyperparameters such as tau, p, kernel width, and ANN architecture, plus assumptions about observability, sampling, and stationarity. No new particles, forces, or conserved quantities are introduced.

free parameters (4)
  • Delay time tau per application = LV: screened over 0.001 to 200 time units; Villin: 200 ps to 10 us; VOO: 2 s, 1 min, 1 h
    Delay times are chosen by hand or by prior knowledge of characteristic time scales and materially control reconstruction quality; Takens' theorem is silent on the choice of tau.
  • Delay embedding dimensionality p = 11 for LV and Villin, 3 for VOO
    Set from Cao E1(d) intrinsic-dimensionality estimates, with p = 2k+1; these choices affect the quality of the learned manifolds.
  • Diffusion map kernel width and manifold dimensionality = Not specified numerically
    Chosen by recursive search and the L-method spectral gap; concrete values are not reported in the main text.
  • ANN architecture and training hyperparameters = Not specified in main text
    Deferred to the Supporting Information; network depth, width, regularization, and epochs are needed to reproduce TAR.
assumptions (5)
  • domain assumption Takens' delay embedding theorem conditions hold for each application, including generic observables, sufficient sampling, p > 2k, and no temporal aliasing.
    Invoked in Sections II through IV; if observables are nongeneric, such as the Villin head-to-tail distance, reconstruction is only guaranteed up to observable symmetries.
  • domain assumption Time-aligned delay vectors correspond one-to-one with instantaneous full states r(t) for training data.
    Section IV.C uses temporal alignment to pair delay vectors with full states; this is a data-construction premise, not a consequence of Takens' theorem.
  • domain assumption Diffusion maps and feedforward ANNs can sufficiently approximate the intrinsic manifold and the M' to M diffeomorphism from finite sampled data.
    The entire TAR pipeline depends on these approximations; the paper provides empirical support but no formal guarantees.
  • domain assumption The low-dimensional manifold learned from the training half of VOO remains representative of test-half dynamics.
    Section VII forward-evaluates TAR on the second half of January 2024; nonstationarity would break the distribution-matching claim.
  • domain assumption Hurst exponents computed with the aggregated-variance estimator identify the claimed regimes without confidence intervals.
    Section VII and Fig. 4(A) report H = 0.40, 0.50, and 0.54 to 0.58 with no uncertainty quantification or multiple-estimator check.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Data-driven reconstruction of dynamical systems using Takens' Theorem, manifold learning, and universal function approximators." pith.science (2026). https://pith.science/paper/C6BY7C44

@misc{pith2026260805477,
  author       = {Pith},
  title        = {Pith review of: Data-driven reconstruction of dynamical systems using Takens' Theorem, manifold learning, and universal function approximators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C6BY7C44}},
  note         = {Machine review of arXiv:2608.05477}
}
read the original abstract

Embedding theorems can be used to provide theoretical guarantees about the relation between low-dimensional observations of a system and its full-dimensional state and dynamics. Such theorems do not, however, provide guidance on observable choice, embedding construction, or methodologies to learn the mapping between the embedding and full-dimensional state. In this work, we develop an algorithmic framework, TAkens Reconstruction (TAR), to analyze and reconstruct arbitrary dynamical systems from low-dimensional time series using an integration of Takens' Delay Embedding Theorem, manifold learning techniques, and universal function approximators. We validate TAR in applications to a variety of simulated and observed dynamical systems and use it to investigate how delay vector structure impacts reconstruction accuracy. In an ecological system, we show that simple predator-prey dynamics can be reconstructed with observations taken over a wide variety of embedding time scales. In molecular dynamics simulations of the protein Villin, we demonstrate how including multiple time delays of the same observable series can be used to improve reconstruction of systems with multiple characteristic time scales. In the trade record of Vanguard S&P 500, we show how the approach exposes underlying dynamical phenomenologies in the data and accurate return predictions over short time horizons without access to full-dimensional market observations. We develop and release an open-source software package to enable the application of TAR to arbitrary dynamical systems.

Figures

Figures reproduced from arXiv: 2608.05477 by the authors.

Figure 1
Figure 1. FIG. 1. Schematic illustration of the TAkens’ Reconstruction (TAR) approach. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. TAR reconstruction of the Lotka-Volterra predator-prey [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Multi-temporal TAR applied to the protein Villin. [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4. TAR analysis of the Vanguard S&P 500 ETF (VOO) returns from January 2024. [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

89 extracted references · 75 canonical work pages

  1. [1]

    S. A. Hollingsworth and R. O. Dror, Molecular dynamics sim- ulation for all, Neuron99, 1129 (2018)

    2018

  2. [2]

    P. G. Wu and L. Brand, Resonance energy transfer: Methods and applications, Analytical Biochemistry218, 1 (1994)

    1994

  3. [3]

    R. Roy, S. Hohng, and T. Ha, A practical guide to single- molecule FRET, Nature Methods5, 507 (2008)

    2008

  4. [4]

    E. R. Deyle and G. Sugihara, Generalized theorems for nonlin- ear state space reconstruction, PLOS ONE6, 1 (2011)

    2011

  5. [5]

    P. Y . Lu, J. Ariño Bernad, and M. Soljaˇci´c, Discovering sparse interpretable dynamics from partial observations, Communica- tions Physics5, 206 (2022)

    2022

  6. [6]

    Berry and S

    T. Berry and S. Das, Learning theory for dynamical systems, SIAM Journal on Applied Dynamical Systems22, 2082 (2023)

    2023

  7. [7]

    O. Yair, R. Talmon, R. R. Coifman, and I. G. Kevrekidis, Re- construction of normal forms by learning informed observation geometries from data, Proceedings of the National Academy of Sciences of the United States of America114, E7865 (2017)

    2017

  8. [8]

    Takens, Detecting strange attractors in turbulence, inDy- namical Systems and Turbulence, Warwick 1980, Lecture Notes in Mathematics, V ol

    F. Takens, Detecting strange attractors in turbulence, inDy- namical Systems and Turbulence, Warwick 1980, Lecture Notes in Mathematics, V ol. 898, edited by D. A. Rand and L.-S. Young (Springer Berlin Heidelberg, Berlin, Heidelberg, 1981) pp. 366–381

    1980

Show all 89 references
  1. [9]

    L. Cao, A. Mees, and K. Judd, Dynamics from multivariate time series, Physica D: Nonlinear Phenomena121, 75 (1998)

    1998

  2. [10]

    Stark, Delay embeddings for forced systems

    J. Stark, Delay embeddings for forced systems. I. Deterministic forcing, Journal of Nonlinear Science9, 255 (1999)

    1999

  3. [11]

    D. S. Broomhead and G. P. King, Extracting qualitative dynam- ics from experimental data, Physica D: Nonlinear Phenomena 20, 217 (1986)

    1986

  4. [12]

    Stark, D

    J. Stark, D. S. Broomhead, M. E. Davies, and J. Huke, Delay embeddings for forced systems. II. Stochastic forcing, Journal of Nonlinear Science13, 519 (2003)

    2003

  5. [13]

    Bara ´nski, Y

    K. Bara ´nski, Y . Gutman, and A.´Spiewak, A probabilistic takens theorem, Nonlinearity33, 4940 (2020)

    2020

  6. [14]

    M. H. Hassoun,Fundamentals of Artificial Neural Networks (MIT Press, Cambridge, MA, 1995)

    1995

  7. [15]

    Wang and A

    J. Wang and A. L. Ferguson, Nonlinear reconstruction of single- molecule free energy surfaces from univariate time series, Phys. Rev. E93, 032412 (2016)

    2016

  8. [16]

    Wang and A

    J. Wang and A. L. Ferguson, Recovery of protein folding fun- nels from single-molecule time series by delay embeddings and manifold learning, J. Phys. Chem. B122, 11931 (2018)

    2018

  9. [17]

    Topel and A

    M. Topel and A. L. Ferguson, Reconstruction of protein struc- tures from single-molecule time series, The Journal of Chemi- cal Physics153, 194102 (2020)

    2020

  10. [18]

    Topel, A

    M. Topel, A. Ejaz, A. Squires, and A. L. Ferguson, Learned reconstruction of protein folding trajectories from noisy single- molecule time series, Journal of Chemical Theory and Compu- tation19, 4654 (2023)

    2023

  11. [19]

    Topel, Kolmogorov-Sinai entropies identify optimal observ- ables for prediction and dynamics reconstruction in chaotic sys- tems, arXiv preprint arXiv:2604.27412 (2026)

    M. Topel, Kolmogorov-Sinai entropies identify optimal observ- ables for prediction and dynamics reconstruction in chaotic sys- tems, arXiv preprint arXiv:2604.27412 (2026)

    2026 arXiv

  12. [20]

    P. J. Wangersky, Lotka-V olterra population models, Annual Re- view of Ecology and Systematics9, 189 (1978)

    1978

  13. [21]

    Brin and G

    M. Brin and G. Stuck,Introduction to Dynamical Systems (Cambridge University Press, 2002)

    2002

  14. [22]

    S. H. Strogatz,Nonlinear Dynamics and Chaos: With Appli- cations to Physics, Biology, Chemistry, and Engineering(CRC Press, 2024)

    2024

  15. [23]

    J. P. Eckmann and D. Ruelle, Ergodic theory of chaos and strange attractors, Reviews of Modern Physics57, 617 (1985)

    1985

  16. [24]

    Sauer, J

    T. Sauer, J. A. Yorke, and M. Casdagli, Embedology, Journal of Statistical Physics65, 579 (1991)

    1991

  17. [25]

    J. B. Tenenbaum, V . de Silva, and J. C. Langford, A global geo- metric framework for nonlinear dimensionality reduction, Sci- ence290, 2319 (2000)

    2000

  18. [26]

    S. T. Roweis and L. K. Saul, Nonlinear dimensionality reduc- tion by locally linear embedding, Science290, 2323 (2000)

    2000

  19. [27]

    R. R. Coifman and S. Lafon, Diffusion maps, Applied and Computational Harmonic Analysis21, 5 (2006)

    2006

  20. [28]

    R. R. Coifman, S. Lafon, A. B. Lee, M. Maggioni, B. Nadler, F. Warner, and S. W. Zucker, Geometric diffusions as a tool for harmonic analysis and structure definition of data: Diffusion maps, Proceedings of the National Academy of Sciences of the United States of America102, 7...

    2005

  21. [29]

    A. E. García, Large-amplitude nonlinear motions in proteins, Physical Review Letters68, 2696 (1992)

    1992

  22. [30]

    Amadei, A

    A. Amadei, A. B. Linssen, and H. J. Berendsen, Essential dy- namics of proteins, Proteins: Structure, Function, and Bioinfor- matics17, 412 (1993)

    1993

  23. [31]

    Hegger, A

    R. Hegger, A. Altis, P. H. Nguyen, and G. Stock, How complex is the dynamics of peptide folding?, Physical Review Letters 98, 028102 (2007)

    2007

  24. [32]

    P. I. Zhuravlev, C. K. Materese, and G. A. Papoian, Decon- structing the native state: energy landscapes, function, and dy- namics of globular proteins, The Journal of Physical Chemistry B113, 8800 (2009)

    2009

  25. [33]

    P. Das, M. Moll, H. Stamati, L. E. Kavraki, and C. Clementi, Low-dimensional, free-energy landscapes of protein-folding re- actions by nonlinear dimensionality reduction, Proceedings of the National Academy of Sciences of the United States of America103, 9885 (2006)

    2006

  26. [34]

    A. L. Ferguson, A. Z. Panagiotopoulos, P. G. Debenedetti, and I. G. Kevrekidis, Systematic determination of order parameters for chain dynamics using diffusion maps, Proceedings of the National Academy of Sciences of the United States of America 107, 13597 (2010)

    2010

  27. [35]

    Whitney, Differentiable manifolds, Annals of Mathematics 37, 645 (1936)

    H. Whitney, Differentiable manifolds, Annals of Mathematics 37, 645 (1936)

    1936

  28. [36]

    M. W. Hirsch,Differential Topology, Graduate Texts in Mathe- matics No. 33 (Springer-Verlag, 1994)

    1994

  29. [37]

    N. H. Packard, J. P. Crutchfield, J. D. Farmer, and R. S. Shaw, Geometry from a time series, Physical Review Letters45, 712 (1980)

    1980

  30. [38]

    Kantz and T

    H. Kantz and T. Schreiber,Nonlinear Time Series Analysis (Cambridge University Press, 2004)

    2004

  31. [39]

    Martin, C

    R. Martin, C. Greve, C. Huerta, A. Wong, J. Koo, and D. Eck- hardt, A robust time-delay selection criterion applied to con- vergent cross mapping, Chaos: An Interdisciplinary Journal of Nonlinear Science34, 093110 (2024)

    2024

  32. [40]

    Scholz, M

    M. Scholz, M. Fraunholz, and J. Selbig, Nonlinear principal component analysis: Neural network models and applications, inPrincipal Manifolds for Data Visualization and Dimension Reduction, Lecture Notes in Computational Science and Engi- neering, V ol. 58 (Springer Berlin Hei...

    2008

  33. [41]

    Cao, Practical method for determining the minimum embed- ding dimension of a scalar time series, Physica D: Nonlinear Phenomena110, 43 (1997)

    L. Cao, Practical method for determining the minimum embed- ding dimension of a scalar time series, Physica D: Nonlinear Phenomena110, 43 (1997)

    1997

  34. [42]

    Nadler, S

    B. Nadler, S. Lafon, R. R. Coifman, and I. G. Kevrekidis, Dif- fusion maps, spectral clustering and eigenfunctions of Fokker- Planck operators, inAdvances in Neural Information Process- ing Systems 18: Proceedings of the 2005 Conference (Neural 17 Information Processing)(The M...

    2005

  35. [43]

    A. L. Ferguson, A. Z. Panagiotopoulos, I. G. Kevrekidis, and P. G. Debenedetti, Nonlinear dimensionality reduction in molecular simulation: The diffusion map approach, Chemical Physics Letters509, 1 (2011)

    2011

  36. [44]

    Nadler, S

    B. Nadler, S. Lafon, R. R. Coifman, and I. G. Kevrekidis, Dif- fusion maps, spectral clustering and reaction coordinates of dy- namical systems, Applied and Computational Harmonic Anal- ysis21, 113 (2006)

    2006

  37. [45]

    A. M. Fraser and H. L. Swinney, Independent coordinates for strange attractors from mutual information, Physical Review A 33, 1134 (1986)

    1986

  38. [46]

    R. R. Coifman, I. G. Kevrekidis, S. Lafon, M. Maggioni, and B. Nadler, Diffusion maps, reduction coordinates, and low dimensional representation of stochastic systems, Multiscale Modeling & Simulation7, 842 (2008)

    2008

  39. [47]

    A. W. Long and A. L. Ferguson, Landmark diffusion maps (L- dMaps): Accelerated manifold learning out-of-sample exten- sion, Applied and Computational Harmonic Analysis47, 190 (2019)

    2019

  40. [48]

    Salvador and P

    S. Salvador and P. Chan, Determining the number of clus- ters/segments in hierarchical clustering/segmentation algo- rithms, in16th IEEE International Conference on Tools with Artificial Intelligence (ICTAI), 2004(2004) pp. 576–584

    2004

  41. [49]

    Williams and M

    C. Williams and M. Seeger, Using the Nyström method to speed up kernel machines, inAdvances in Neural Information Pro- cessing Systems, V ol. 13 (MIT Press, 2000) pp. 682–688

    2000

  42. [50]

    Wang and A

    J. Wang and A. L. Ferguson, A study of the morphol- ogy, dynamics, and folding pathways of ring polymers with supramolecular topological constraints using molecular simula- tion and nonlinear manifold learning, Macromolecules51, 598 (2018)

    2018

  43. [51]

    I. G. Kevrekidis, C. W. Gear, and G. Hummer, Equation-free modeling for complex systems, inHandbook of Materials Mod- eling: Methods(Springer Netherlands, Dordrecht, 2005) pp. 1453–1475

    2005

  44. [52]

    I. G. Kevrekidis and G. Samaey, Equation-free multiscale com- putation: Algorithms and applications, Annual Review of Phys- ical Chemistry60, 321 (2009)

    2009

  45. [53]

    M. S. Jones, K. Shmilovich, and A. L. Ferguson, DiAMoND- Back: Diffusion-denoising autoregressive model for non- deterministic backmapping of Cαprotein traces, Journal of Chemical Theory and Computation19, 7908 (2023)

    2023

  46. [54]

    M. S. Jones, S. Khanna, and A. L. Ferguson, FlowBack: A gen- eralized flow-matching approach for biomolecular backmap- ping, Journal of Chemical Information and Modeling65, 672 (2025)

    2025

  47. [55]

    Sidky, W

    H. Sidky, W. Chen, and A. L. Ferguson, Molecular latent space simulators, Chemical Science11, 9459 (2020)

    2020

  48. [56]

    H. Ye, R. J. Beamish, S. M. Glaser, S. C. H. Grant, C.-H. Hsieh, L. J. Richards, J. T. Schnute, and G. Sugihara, Equation-free mechanistic ecosystem forecasting using empirical dynamic modeling, Proceedings of the National Academy of Sciences of the United States of America112...

    2015

  49. [57]

    Scheffer, J

    M. Scheffer, J. Bascompte, W. A. Brock, V . Brovkin, S. R. Car- penter, V . Dakos, H. Held, E. H. van Nes, M. Rietkerk, and G. Sugihara, Early-warning signals for critical transitions, Na- ture461, 53 (2009)

    2009

  50. [58]

    J. C. Butcher, On Runge-Kutta processes of high order, Journal of the Australian Mathematical Society4, 179 (1964)

    1964

  51. [59]

    Haken,Synergetics: An Introduction(Springer Berlin Hei- delberg, 1983)

    H. Haken,Synergetics: An Introduction(Springer Berlin Hei- delberg, 1983)

    1983

  52. [60]

    C. D. Brummitt, K.-M. Lee, and K.-I. Goh, Multiplexity- facilitated cascades in networks, Phys. Rev. E85, 045102 (2012)

    2012

  53. [61]

    Easley and J

    D. Easley and J. Kleinberg, Information cascades, inNetworks, Crowds, and Markets: Reasoning about a Highly Connected World(Cambridge University Press, 2010) Chap. 16, pp. 483– 508

    2010

  54. [62]

    Junghare, S

    V . Junghare, S. Bhattacharya, K. Ansari, and S. Hazra, Markov state models of molecular simulations to study protein folding and dynamics, inProtein Folding Dynamics and Stability: Ex- perimental and Computational Methods(Springer Nature Sin- gapore, 2023) pp. 147–164

    2023

  55. [63]

    Henzler-Wildman and D

    K. Henzler-Wildman and D. Kern, Dynamic personalities of proteins, Nature450, 964 (2007)

    2007

  56. [64]

    Lindorff-Larsen, S

    K. Lindorff-Larsen, S. Piana, R. O. Dror, and D. E. Shaw, How fast-folding proteins fold, Science334, 517 (2011)

    2011

  57. [65]

    Humphrey, A

    W. Humphrey, A. Dalke, and K. Schulten, VMD: Visual molec- ular dynamics, Journal of Molecular Graphics14, 33 (1996)

    1996

  58. [66]

    Hotelling, Relations between two sets of variates, Biometrika28, 321 (1936)

    H. Hotelling, Relations between two sets of variates, Biometrika28, 321 (1936)

    1936

  59. [67]

    Madhavan, Market microstructure: A survey, Journal of Fi- nancial Markets3, 205 (2000)

    A. Madhavan, Market microstructure: A survey, Journal of Fi- nancial Markets3, 205 (2000)

    2000

  60. [68]

    Bouchaud, J

    J.-P. Bouchaud, J. Bonart, J. Donier, and M. Gould,Trades, Quotes and Prices: Financial Markets Under the Microscope (Cambridge University Press, 2018)

    2018

  61. [69]

    J. Y . Campbell, A. W. Lo, and A. C. MacKinlay,The Economet- rics of Financial Markets(Princeton University Press, Prince- ton, NJ, 1997)

    1997

  62. [70]

    O’Hara,Market Microstructure Theory(Blackwell Publish- ers, 1995)

    M. O’Hara,Market Microstructure Theory(Blackwell Publish- ers, 1995)

    1995

  63. [71]

    Roll, A simple implicit measure of the effective bid–ask spread in an efficient market, The Journal of Finance39, 1127 (1984)

    R. Roll, A simple implicit measure of the effective bid–ask spread in an efficient market, The Journal of Finance39, 1127 (1984)

    1984

  64. [72]

    O. E. Barndorff-Nielsen, Normal inverse Gaussian distributions and stochastic volatility modelling, Scandinavian Journal of Statistics24, 1 (1997)

    1997

  65. [73]

    P. K. Clark, A subordinated stochastic process model with finite variance for speculative prices, Econometrica41, 135 (1973)

    1973

  66. [74]

    E. F. Fama, Random walks in stock market prices, Financial Analysts Journal51, 75 (1995)

    1995

  67. [75]

    E. F. Fama, The behavior of stock-market prices, The Journal of Business38, 34 (1965)

    1965

  68. [76]

    Mandelbrot, The variation of certain speculative prices, The Journal of Business36, 394 (1963)

    B. Mandelbrot, The variation of certain speculative prices, The Journal of Business36, 394 (1963)

    1963

  69. [77]

    R. F. Engle, Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation, Econo- metrica50, 987 (1982)

    1982

  70. [78]

    Bollerslev, Generalized autoregressive conditional het- eroskedasticity, Journal of Econometrics31, 307 (1986)

    T. Bollerslev, Generalized autoregressive conditional het- eroskedasticity, Journal of Econometrics31, 307 (1986)

    1986

  71. [79]

    Bollerslev, A conditionally heteroskedastic time series model for speculative prices and rates of return, The Review of Eco- nomics and Statistics69, 542 (1987)

    T. Bollerslev, A conditionally heteroskedastic time series model for speculative prices and rates of return, The Review of Eco- nomics and Statistics69, 542 (1987)

    1987

  72. [80]

    S. R. Baker, N. Bloom, S. J. Davis, K. Kost, M. Sammon, and T. Viratyosin, The unprecedented stock market reaction to COVID-19, The Review of Asset Pricing Studies10, 742 (2020)

    2020

  73. [81]

    Wharton Research Data Services, Executed trades data for vanguard S&P 500 ETF (VOO) (2024), dataset re- trieved from Wharton Research Data Services (https://wrds- www.wharton.upenn.edu/); all executed sales transactions dur- ing trading hours in January 2024

    2024

  74. [82]

    P. R. Hansen and A. Lunde, Realized variance and market mi- crostructure noise, Journal of Business & Economic Statistics 24, 127 (2006)

    2006

  75. [83]

    M. S. Taqqu, V . Teverovsky, and W. Willinger, Estimators for long-range dependence: An empirical study, Fractals3, 785 18 (1995)

    1995

  76. [84]

    H. E. Hurst, The problem of long-term storage in reservoirs, International Association of Scientific Hydrology. Bulletin1, 13 (1956)

    1956

  77. [85]

    H. E. Hurst, A suggested statistical model of some time series which occur in nature, Nature180, 494 (1957)

    1957

  78. [86]

    Weron, Estimating long-range dependence: Finite sample properties and confidence intervals, Physica A: Statistical Me- chanics and its Applications312, 285 (2002)

    R. Weron, Estimating long-range dependence: Finite sample properties and confidence intervals, Physica A: Statistical Me- chanics and its Applications312, 285 (2002)

    2002

  79. [87]

    Aris, How to get the most out of an equation without really trying, Chemical Engineering Education10, 114 (1976)

    R. Aris, How to get the most out of an equation without really trying, Chemical Engineering Education10, 114 (1976)

    1976

  80. [88]

    Ma and A

    A. Ma and A. R. Dinner, Automatic method for identifying reaction coordinates in complex systems, Journal of Physical Chemistry B109, 6769 (2005)

    2005

  81. [89]

    F. P. Kemeth, S. W. Haugland, F. Dietrich, T. Bertalan, Q. Li, E. M. Bollt, R. Talmon, K. Krischer, and I. G. Kevrekidis, An equal space for complex data with unknown internal order: Observability, gauge invariance and manifold learning, arXiv preprint arXiv.1708.05406 (2017)

    2017 arXiv

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.