Pith. sign in

REVIEW 3 major objections 6 minor 39 references

A Mathematical Framework for Topological Causal Data Analysis

T0 review · 3 major / 6 minor · reviewed 2026-07-31 · grok-4.5

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

Pith's one-line read Causal effects on shapes and structured outcomes become well-defined once topology is applied only after interventions and assumptions are fixed, and outcome-level averaging generally differs from law-level topology.

desk verdict Solid architecture paper that cleanly unifies concurrent TDA-causal work and proves a few real transfer theorems; worth engaging as framework, not as a new estimator suite. read the letter →

arxiv 2607.28161 v1 pith:K7FSIN2K submitted 2026-07-30 stat.ME math.STstat.MLstat.TH

Hugo Gobato Souto , Ioannis Diamantis This is my paper · ORCID
classification stat.MEmath.STstat.MLstat.TH MSC 62D2055N3162G0562R40
keywords topologicalcausaldataanalysispersistenthomologypotentialoutcomesdistribution-leveleffectsBanach-valuedtreatmentignorabilitystabilitytransferg-formula
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

Many modern outcomes—tumour shapes, brain networks, point clouds, climate fields—are not numbers, so the usual treatment-minus-control difference is undefined or scientifically empty. This paper introduces Topological Causal Data Analysis (TCDA): a four-layer setup that keeps the observation space, the causal model, the topological summary, and the causal question strictly separate. Topology never defines the intervention; it only supplies a stable shape-sensitive feature after the causal target is chosen. The framework splits into two levels that usually disagree. Outcome-level TCDA maps each potential outcome into a Banach space (for example a persistence landscape) and takes the average treatment effect there; distribution-level TCDA first forms the interventional laws and then applies topology to those laws, catching changes such as one cluster splitting into two even when means stay the same. The paper proves when the two contrasts agree, gives identification and doubly robust formulas for Banach-valued effects, transfers Lipschitz stability of topological maps into error bounds on the causal contrasts, and shows that observational topology can separate only restricted mechanism classes—it cannot by itself recover causal structure.

What carries the argument

The TCDA problem tuple PTCDA = (S, M(G), T, C), which forces separate specification of observation space, causal-model class, topological representation, and causal query; together with the affine-mean functional A and the characterization that outcome- and distribution-level contrasts agree everywhere exactly when T_dist − A is constant.

What would settle it

Construct two interventional laws with the same mean outcome-level topological summary but different distribution-level topology (for example one versus two density clusters), estimate both contrasts from data generated under known exchangeability, and check whether the empirical contrasts match the paper’s non-commutation prediction and whether the plug-in error tracks the stated Wasserstein or bottleneck bounds.

Watch

Extended reading notes

Core claim

Under the four-layer TCDA problem, outcome-level Banach-valued topological average treatment effects are identified by the ordinary standardization, inverse-probability, and augmented formulas; distribution-level targets are identified by the g-formula applied to interventional laws before topology; and the two level contrasts agree for every pair of laws if and only if the distribution-level map differs from the mean functional by a constant. Lipschitz topology then transfers directly to stability and plug-in bounds on the causal contrasts.

Load-bearing premise

Identification still requires that treatment assignment is independent of potential outcomes given covariates (or the weaker, untestable, representation-specific topological ignorability), plus positivity; if that fails, none of the causal contrasts are identified from observational data.

Editorial extensions

If this is right

  • Shape-, image-, and network-valued treatment effects can be stated and identified without forcing the outcome into a single scalar.
  • A zero classical average treatment effect need not imply zero topological effect once topology is applied to the interventional laws.
  • Doubly robust and cross-fitted estimators extend, under product-rate conditions, to Banach-valued persistence summaries such as silhouettes and landscapes.
  • Observational persistent homology can at best separate restricted mechanism classes with a positive margin; it cannot orient edges or replace conditional independence or intervention assumptions.
  • Target-specific topological ignorability can identify a coarse covariate-standardized effect without identifying full interventional laws, but only inside the fiber of the chosen non-injective summary.

Reading between the lines

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

  • Clinical and imaging trials that currently collapse tumours or organs to volume or a few landmarks could re-analyze the same scans under outcome-level TCDA to recover multiscale shape effects that scalar endpoints miss.
  • The non-commutation result suggests a practical diagnostic: if outcome-level and distribution-level estimates diverge sharply, treatment is rearranging population geometry rather than shifting typical individuals.
  • Stability-transfer bounds give a concrete design rule for choosing filtrations and mass parameters so that estimation error in the interventional laws stays below a scientifically meaningful topological threshold.
  • Topology-assisted discovery will remain limited to low-dimensional additive-noise or latent-geometry settings unless paired with independent-noise or invariance assumptions the paper deliberately excludes.
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

3 major / 6 minor

Summary. The paper proposes Topological Causal Data Analysis (TCDA) as a four-layer architecture PTCDA=(S,M(G),T,C) that keeps observation space, causal-model class, topological representation, and causal query separate. It distinguishes outcome-level effects (Banach-valued maps of individual potential outcomes, with standardization/IPW/augmented identification and product-rate remainders) from distribution-level effects (topology applied to interventional laws via the g-formula), and characterizes agreement of the two contrasts by constancy of T_dist-A (Theorem 5.9). Lipschitz stability of filtrations and vectorizations is transferred to causal contrasts and plug-in estimators; target-specific topological ignorability and the limited role of observational topology in discovery are placed inside the same framework, with explicit attribution to concurrent outcome-level and ignorability results.

Significance. If the framework is adopted, it gives a clean vocabulary for causal questions about shapes, images, networks, and spatial fields where Y^1-Y^0 is undefined, and it prevents conflating intervention, identification, and topological feature choice. The main technical payoffs that stand on their own are the non-commutation/agreement characterization (Theorem 5.9), the metric-matched stability-transfer and plug-in bounds (Section 6, especially DTM/W_2), and the precise delimitation of discovery and topological ignorability. Strengths include careful attribution to Kim–Lee, Saki–Faghihi, and Shin et al., explicit non-claims (no generic robustness to hidden confounding; topology alone does not orient edges), and correctly matched diagram metrics (Lemma 6.7, Table 1). The contribution is architectural and clarifying rather than a new identification principle or a full inferential theory for general Banach-valued summaries.

major comments (3)
  1. [§4.2–4.3] §4.2–4.3 and Contribution 2: Proposition 4.3 gives the standard doubly robust remainder in a Banach space, but the manuscript correctly notes that this does not yield a CLT. Functional inference is imported only for power-weighted silhouettes (Kim–Lee). As written, the claim to “formulate identification and doubly robust representations for Banach-space-valued summaries” is accurate for population identities, yet readers may over-read it as delivering usable inference for landscapes, images, or Betti curves. Please state explicitly in the contribution list and at the end of §4.2 which objects have complete estimation theory in this paper versus which only inherit population DR identities, and avoid language that suggests general root-n Banach inference is established here.
  2. [§5.3, Prop. 6.15, §6.6] §5.3 and Proposition 6.15: Plug-in consistency and rate transfer are conditional on d_P(P̂^a, P^a_Y)→0 (and W_2 for DTM). The paper does not prove that Hájek, g-formula, or projected estimators achieve those metrics under the stated positivity conditions alone, and it notes this at the end of §6.6. That caveat should be elevated next to Corollary 5.4 and Proposition 6.15 (e.g., a short remark that rate results are transfer principles, not end-to-end estimator theorems), so that distribution-level “plug-in consistency” is not mistaken for a free statistical guarantee.
  3. [§1, §4.3, §8] Dependence on concurrent preprints: Large parts of the outcome-level inferential story (§4.3) and the topological-ignorability material (§8, Proposition 8.2) are attributed restatements of Kim–Lee and Saki–Faghihi. The independent core (architecture, Theorem 5.9, stability organization, discovery limits) is real but narrower. Please add a short “relation to concurrent work” subsection that itemizes, theorem-by-theorem, what is proved here versus what is cited, so the paper’s incremental contribution is auditable if those preprints change.
minor comments (6)
  1. [§2.1] Notation for interventional laws switches among P^a_Y, L(Y^a), and P^a_{Y,M}. A single convention in §2.1 would reduce friction.
  2. [§6, Table 1] Table 1 is helpful; add a one-line pointer in the caption to the propositions that instantiate each row (6.10, 6.12, 6.14).
  3. [§5.4–5.5, §9] Example 5.10–5.11 and §9 figures are conceptual only. Even a small simulated numerical check (e.g., estimated Δ_dist under known P^a_Y) would make the non-commutation message more concrete without turning the paper empirical.
  4. [§2.3, §5.6] Assumption 2.5 notes that finite second moment plus Lipschitz DTM does not imply q-tameness; consider flagging this again in §5.6 where DTM effects are defined as estimands.
  5. Minor typos and spacing artifacts appear in the compiled text (e.g., split words such as “topolog-ical”, “g-formula” line breaks). A proofreading pass is needed.
  6. [§7] Definition 7.1–7.2 and Proposition 7.5 are clear; the support-based Example 7.3 could cite the Remark 7.6 stability warning in the example statement itself so readers do not take β_1(supp) as a recommended estimator.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TCDA is a definitional framework plus standard identification/stability transfer under external causal assumptions.

full rationale

The paper defines outcome-level and distribution-level targets from potential outcomes and interventional laws, then identifies them under the usual external assumptions (consistency, exchangeability/positivity, or the weaker attributed topological ignorability). Theorem 5.9’s agreement criterion (T_dist−A constant) is an algebraic characterization of when two defined contrasts coincide, not a prediction forced by fitting or by smuggling the conclusion into the premises. Stability theorems transfer Lipschitz constants of the chosen representation to causal contrasts by triangle/Bochner–Jensen arguments; they do not redefine the estimands as their own bounds. Concurrent Kim–Lee and Saki–Faghihi results are cited with explicit attribution for silhouette inference and topological ignorability rather than silently re-derived as first principles of the present authors. There is no fitted parameter re-labeled as a prediction, no self-citation uniqueness theorem forbidding alternatives, and no load-bearing self-definitional loop. Score 0 is appropriate.

Assumptions & free parameters 4 free parameters · 8 assumptions · 3 invented entities

The framework inherits classical causal identification axioms and standard TDA regularity, then adds modeling choices (filtration, degree, vectorization, DTM mass m) as part of the estimand. No physical constants are fitted. Invented items are definitional packaging (TCDA/PTCDA and named estimands), not new latent forces. Load-bearing uncertainty sits in untestable exchangeability/topological ignorability and in q-tameness/Lipschitz hypotheses on the chosen pipeline.

free parameters (4)
  • Filtration F and homological degree k
    Scientific modeling choice that defines the estimand; different filtrations (VR, cubical, sublevel, DTM) yield different causal targets. Not fitted to data in the paper, but free in applications.
  • DTM mass parameter m ∈ (0,1)
    Enters both the estimand and the stability constant m^{-1/2}; smaller m keeps local structure but worsens constants (§5.6, Prop 6.14).
  • Vectorization Φ and diagram metric (d_B vs W_p)
    Chooses codomain Banach/metric structure and Lipschitz constant κ_Φ; must match diagram metric (Def 6.6, Lemma 6.7).
  • Propensity truncation floor ε and cross-fit folds K
    Estimation hyperparameters controlling weight boundedness and product-rate bias in §4.2; not numerically tuned here.
assumptions (8)
  • domain assumption Consistency: Y = Y^A a.s. (Assumption 2.1)
    Standard potential-outcome link; required for all identification theorems.
  • domain assumption Conditional exchangeability (Y^0,Y^1) ⊥ A | X, or arm-wise weak exchangeability, or conditional topological ignorability for T_dist (Assumptions 2.2, Def 8.1)
    Untestable causal core; topological ignorability can be strictly weaker only with non-injective T_dist and distinct laws in one fiber.
  • domain assumption Strict/strong positivity of propensity e(X) (Assumptions 2.3–2.4)
    Needed for IPW/g-formula identification and bounded weights in DR estimation.
  • domain assumption q-tameness of persistence modules for outcomes and laws used (Assumption 2.5)
    Guarantees diagrams exist; automatic for finite complexes, not free for population DTM/sublevel modules.
  • standard math Bochner integrability E∥T_out(Y^a)∥_B < ∞ and separability of Banach target B
    Required to define E[Z^a] and linear contrasts in outcome-level TCDA (§2.3, Def 3.3).
  • domain assumption Lipschitz stability of T_out / T_dist in the matched outcome or law metric (Assumptions 6.1, 6.4)
    Transfers to causal-contrast and plug-in bounds; fails if vectorization/diagram metric are mismatched.
  • standard math Classical bottleneck/VR/DTM stability theorems (Cohen–Steiner et al., Chazal et al.)
    Imported as Lemmas 6.9, 6.11, 6.13 without reproof; underwrite Table 1 constants.
  • standard math Standard Borel observation spaces so regular conditional laws exist
    Stated in §2.1 to justify kernels Q_a and g-formula integrals.
invented entities (3)
  • TCDA problem tuple PTCDA = (S, M(G), T, C)
    purpose: Force separate specification of observation space, causal-model class, topological representation, and causal query.
    Definitional packaging (Def 3.1); organizes prior constructions rather than positing a new physical object.
  • Outcome-level TATE_out and distribution-level Δ_dist / δ_dist / τ_dist
    purpose: Name the two aggregation orders and the covariate-standardized within-stratum topological effect.
    Estimand definitions (Defs 3.3, 5.2, 5.5); empirical content depends on external causal assumptions and chosen T.
  • T_obs-identifiability and topological separation margin η for restricted discovery
    purpose: Formalize when observational topology can distinguish causal equivalence classes on a restricted model class.
    Defs 7.1–7.2 plus Example 7.3; illustrates limits rather than claiming generic discovery power.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Mathematical Framework for Topological Causal Data Analysis." pith.science (2026). https://pith.science/paper/K7FSIN2K

@misc{pith2026260728161,
  author       = {Pith},
  title        = {Pith review of: A Mathematical Framework for Topological Causal Data Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K7FSIN2K}},
  note         = {Machine review of arXiv:2607.28161}
}
abstract

Many modern outcomes, including images, point clouds, networks, and spatial fields, are structured objects for which \(Y^1-Y^0\) may be undefined or scientifically inadequate. We introduce \emph{Topological Causal Data Analysis} (TCDA), a framework separating the observation space, causal-model class, topological representation, and causal query. Topology does not define interventions; it supplies stable, shape-sensitive summaries after causal assumptions have been specified. We distinguish outcome-level TCDA, which transforms individual potential outcomes, from distribution-level TCDA, which transforms interventional outcome laws, and characterize when outcome and distribution level contrasts agree. Building on recent outcome-level theory, we formulate identification and doubly robust representations for Banach-space-valued summaries. At the distribution level, we identify targets through the standard causal \(g\)-formula and derive stability-transfer bounds and plug-in consistency. We also place target-specific topological ignorability within the framework, clarifying when a covariate-standardized coarse effect can be identified without identifying the full interventional laws. Finally, we delimit the role of observational topology in causal discovery: it can assist diagnosis on restricted model classes but cannot by itself identify causal structure.

Figures

Figures reproduced from arXiv: 2607.28161 by the authors.

Figure 1
Figure 1. Causal structure (top) and observational support (bottom) for the three model families [PITH_FULL_IMAGE:figures/full_fig_p028_1.png] view at source ↗
Figure 2
Figure 2. Illustration of a shape-valued treatment effect. The potential outcomes [PITH_FULL_IMAGE:figures/full_fig_p032_2.png] view at source ↗
Figure 3
Figure 3. Residual topology as a model diagnostic. Although [PITH_FULL_IMAGE:figures/full_fig_p033_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Illustration of a distribution-level treatment effect. The two interventional laws have [PITH_FULL_IMAGE:figures/full_fig_p034_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 4 canonical work pages

  1. [1]

    Persis- 34 tence images: A stable vector representation of persistent homology.Journal of Machine Learning Research, 18(8):1–35, 2017

    Henry Adams, Tegan Emerson, Michael Kirby, Rachel Neville, Chris Peterson, Patrick Shipman, Sofya Chepushtanova, Eric Hanson, Francis Motta, and Lori Ziegelmeier. Persis- 34 tence images: A stable vector representation of persistent homology.Journal of Machine Learning Research, 18(8):1–35, 2017. URLhttp://jmlr.org/papers/v18/16-337.html

    2017

  2. [2]

    Cambridge University Press, 2018

    Jean-Daniel Boissonnat, Frédéric Chazal, and Mariette Yvinec.Geometric and Topological Inference. Cambridge University Press, 2018

    2018

  3. [3]

    Statistical topological data analysis using persistence landscapes.J

    Peter Bubenik. Statistical topological data analysis using persistence landscapes.J. Mach. Learn. Res., 16(1):77–102, January 2015. ISSN 1532-4435

    2015

  4. [4]

    A persistence landscapes toolbox for topological statis- tics.Journal of Symbolic Computation, 78:91–114, January 2017

    Peter Bubenik and Paweł Dłotko. A persistence landscapes toolbox for topological statis- tics.Journal of Symbolic Computation, 78:91–114, January 2017. ISSN 0747-7171. doi: 10.1016/j.jsc.2016.03.009. URLhttp://dx.doi.org/10.1016/j.jsc.2016.03.009

    doi:10.1016/j.jsc.2016.03.009 2017

  5. [5]

    Oudot, and Donald R

    Mickaël Buchet, Frédéric Chazal, Steve Y. Oudot, and Donald R. Sheehy. Efficient and robust persistent homology for measures.Computational Geometry, 58:70–96, 2016. doi: 10.1016/j.comgeo.2016.07.001

    doi:10.1016/j.comgeo.2016.07.001 2016

  6. [6]

    Topology and data.Bulletin of the American Mathematical Society, 46 (2):255–308, January 2009

    Gunnar Carlsson. Topology and data.Bulletin of the American Mathematical Society, 46 (2):255–308, January 2009. ISSN 0273-0979. doi: 10.1090/s0273-0979-09-01249-x. URL http://dx.doi.org/10.1090/S0273-0979-09-01249-X

    doi:10.1090/s0273-0979-09-01249-x 2009

  7. [7]

    SlicedWassersteinkernelforpersistence diagrams

    MathieuCarrière, MarcoCuturi, andSteveOudot. SlicedWassersteinkernelforpersistence diagrams. In Doina Precup and Yee Whye Teh, editors,Proceedings of the 34th Interna- tional Conference on Machine Learning, volume 70 ofProceedings of Machine Learning Re- search, pages 664–673. PMLR, 06–11 Aug 2017. URLhttps://proceedings.mlr.press/ v70/carriere17a.html

    2017

  8. [8]

    Persistence stability for geometric com- plexes.Geometriae Dedicata, 173:193–214, 2014

    Frédéric Chazal, Vin de Silva, and Steve Oudot. Persistence stability for geometric com- plexes.Geometriae Dedicata, 173:193–214, 2014. doi: 10.1007/s10711-013-9937-z

    doi:10.1007/s10711-013-9937-z 2014

Show all 39 references
  1. [9]

    Springer, 2016

    Frédéric Chazal, Vin de Silva, Marc Glisse, and Steve Oudot.The Structure and Stability of Persistence Modules. Springer, 2016

    2016

  2. [10]

    An introduction to topological data analysis: Fun- damental and practical aspects for data scientists.Frontiers in Artificial Intelligence, 4,

    Frédéric Chazal and Bertrand Michel. An introduction to topological data analysis: Fun- damental and practical aspects for data scientists.Frontiers in Artificial Intelligence, 4,

  3. [11]

    Geometric inference for probability measures.Foundations of Computational Mathematics, 11(6):733–751, 2011

    Frédéric Chazal, David Cohen-Steiner, and Quentin Mérigot. Geometric inference for probability measures.Foundations of Computational Mathematics, 11(6):733–751, 2011. ISSN 1615-3383. doi: 10.1007/s10208-011-9098-0. URLhttp://dx.doi.org/10.1007/ s10208-011-9098-0

    2011 doi

  4. [12]

    Stochastic convergence of persistence landscapes and silhouettes

    Frédéric Chazal, Brittany Terese Fasy, Fabrizio Lecci, Alessandro Rinaldo, and Larry Wasserman. Stochastic convergence of persistence landscapes and silhouettes. InPro- ceedings of the thirtieth annual symposium on Computational geometry, SOCG’14, page 474–483. ACM, 2014. doi:...

    2014

  5. [13]

    Double/debiased machine learning for treatment and structural parameters.The Econometrics Journal, 21(1):C1–C68, January 2018

    Victor Chernozhukov, Denis Chetverikov, Mert Demirer, Esther Duflo, Christian Hansen, Whitney Newey, and James Robins. Double/debiased machine learning for treatment and structural parameters.The Econometrics Journal, 21(1):C1–C68, January 2018. ISSN 1368-423X. doi: 10.1111/ec...

    2018 doi

  6. [14]

    Stability of persistence diagrams

    David Cohen-Steiner, Herbert Edelsbrunner, and John Harer. Stability of persistence diagrams. InProceedings of the twenty-first annual symposium on Computational geometry, SoCG05, page 263–271. ACM, 2005. doi: 10.1145/1064092.1064133. URLhttp://dx.doi. org/10.1145/1064092.1064133

    2005

  7. [15]

    Lipschitz functions have l p -stable persistence.Foundations of Computational Mathematics, 10(2): 127–139, January 2010

    David Cohen-Steiner, Herbert Edelsbrunner, John Harer, and Yuriy Mileyko. Lipschitz functions have l p -stable persistence.Foundations of Computational Mathematics, 10(2): 127–139, January 2010. ISSN 1615-3383. doi: 10.1007/s10208-010-9060-6. URLhttp: //dx.doi.org/10.1007/s102...

    2010 doi

  8. [16]

    Uhl.Vector Measures

    Joseph Diestel and John J. Uhl.Vector Measures. Number 15 in Mathematical Surveys. American Mathematical Society, Providence, RI, 1977

    1977

  9. [17]

    Ameri- can Mathematical Society, 2010

    Herbert Edelsbrunner and John Harer.Computational Topology: An Introduction. Ameri- can Mathematical Society, 2010

    2010

  10. [18]

    Topological residual asymmetry for bivariate causal direction

    Mouad El Bouchattaoui. Topological residual asymmetry for bivariate causal direction. arXiv preprint arXiv:2602.00427, 2026. URLhttps://arxiv.org/abs/2602.00427

    2026

  11. [19]

    Hernán and James M

    Miguel A. Hernán and James M. Robins.Causal Inference: What If. Chapman & Hall/CRC, 2020

    2020

  12. [20]

    A topological perspective on causal inference

    Duligur Ibeling and Thomas Icard. A topological perspective on causal inference. In Advances in Neural Information Processing Systems, volume 34, pages 5608–5619. Cur- ran Associates, Inc., 2021. URLhttps://proceedings.neurips.cc/paper/2021/hash/ 2c463dfdde588f3bfc60d53118c10d...

    2021

  13. [21]

    Imbens and Donald B

    Guido W. Imbens and Donald B. Rubin.Causal Inference for Statistics, Social, and Biomedical Sciences. Cambridge University Press, 2015

    2015

  14. [22]

    Kennedy.Semiparametric Doubly Robust Targeted Double Machine Learning: A Review, page 207–236

    Edward H. Kennedy.Semiparametric Doubly Robust Targeted Double Machine Learning: A Review, page 207–236. Chapman and Hall/CRC, October 2024. ISBN 9781003216223. doi: 10.1201/9781003216223-10. URLhttp://dx.doi.org/10.1201/9781003216223-10

    2024 doi

  15. [23]

    Topological causal effects.Preprint, arXiv:2603.02289, 2026

    Kwangho Kim and Hajin Lee. Topological causal effects.Preprint, arXiv:2603.02289, 2026. doi: 10.48550/ARXIV.2603.02289. URLhttps://arxiv.org/abs/2603.02289

    2026 doi

  16. [24]

    Kernel method for persistence diagrams via kernel embedding and weight factor.Journal of Machine Learning Research, 18(189):1–41, 2018

    Genki Kusano, Kenji Fukumizu, and Yasuaki Hiraoka. Kernel method for persistence diagrams via kernel embedding and weight factor.Journal of Machine Learning Research, 18(189):1–41, 2018. URLhttp://jmlr.org/papers/v18/17-317.html

    2018

  17. [25]

    Statisti- cal topological data analysis - a kernel perspective

    Roland Kwitt, Stefan Huber, Marc Niethammer, Weili Lin, and Ulrich Bauer. Statisti- cal topological data analysis - a kernel perspective. In C. Cortes, N. Lawrence, D. Lee, 36 M. Sugiyama, and R. Garnett, editors,Advances in Neural Information Processing Sys- tems, volume 28. ...

    2015

  18. [26]

    Persistence fisher kernel: A riemannian manifold kernel for persistence diagrams

    Tam Le and Makoto Yamada. Persistence fisher kernel: A riemannian manifold kernel for persistence diagrams. In S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa- Bianchi, and R. Garnett, editors,Advances in Neural Information Processing Systems, vol- ume 31. Curran Ass...

    2018

  19. [27]

    Probability measures on the space of persistence diagrams.Inverse Problems, 27(12):124007, November 2011

    Yuriy Mileyko, Sayan Mukherjee, and John Harer. Probability measures on the space of persistence diagrams.Inverse Problems, 27(12):124007, November 2011. ISSN 1361-6420. doi: 10.1088/0266-5611/27/12/124007. URLhttp://dx.doi.org/10.1088/0266-5611/ 27/12/124007

    2011 doi

  20. [28]

    Oudot.Persistence Theory: From Quiver Representations to Data Analysis

    Steve Y. Oudot.Persistence Theory: From Quiver Representations to Data Analysis. American Mathematical Society, 2015

    2015

  21. [29]

    Cambridge University Press, 2 edition, 2009

    Judea Pearl.Causality. Cambridge University Press, 2 edition, 2009

    2009

  22. [30]

    James Robins. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect.Mathematical Modelling, 7(9-12):1393–1512, 1986. ISSN 0270-0255. doi: 10.1016/0270-0255(86)90088-6. URLhttp://dx....

    1986 doi

  23. [31]

    Donald B. Rubin. Estimating causal effects of treatments in randomized and nonran- domized studies.Journal of Educational Psychology, 66(5):688–701, October 1974. ISSN 0022-0663. doi: 10.1037/h0037350. URLhttp://dx.doi.org/10.1037/h0037350

    1974 doi

  24. [32]

    Beyond means: Topological causal effects under persistent- homology ignorability.Preprint, arXiv:2603.14169, 2026

    Amir Saki and Usef Faghihi. Beyond means: Topological causal effects under persistent- homology ignorability.Preprint, arXiv:2603.14169, 2026. doi: 10.48550/ARXIV.2603. 14169. URLhttps://arxiv.org/abs/2603.14169

    2026 doi arXiv

  25. [33]

    Absolute average and median treatment effects as causal estimands on metric spaces.Preprint, arXiv:2407.03726,

    Ha-Young Shin, Kyusoon Kim, Kwonsang Lee, and Hee-Seok Oh. Absolute average and median treatment effects as causal estimands on metric spaces.Preprint, arXiv:2407.03726,

    arXiv

  26. [34]

    Wasserstein stability for persistence diagrams, 2020

    Primoz Skraba and Katharine Turner. Wasserstein stability for persistence diagrams, 2020. URLhttps://arxiv.org/abs/2006.16824

    2020 arXiv

  27. [35]

    MIT Press, 2 edition, 2000

    Peter Spirtes, Clark Glymour, and Richard Scheines.Causation, Prediction, and Search. MIT Press, 2 edition, 2000

    2000

  28. [36]

    Fréchet means for distributions of persistence diagrams.Discrete & Computational Geometry, 52(1):44–70,

    Katharine Turner, Yuriy Mileyko, Sayan Mukherjee, and John Harer. Fréchet means for distributions of persistence diagrams.Discrete & Computational Geometry, 52(1):44–70,

  29. [2014]

    doi: 10.1007/s00454-014-9604-7

    ISSN 1432-0444. doi: 10.1007/s00454-014-9604-7. URLhttp://dx.doi.org/10. 1007/s00454-014-9604-7. 37

    doi

  30. [2021]

    doi: 10.3389/frai.2021.667963

    ISSN 2624-8212. doi: 10.3389/frai.2021.667963. URLhttp://dx.doi.org/10.3389/ frai.2021.667963

    2021

  31. [2024]

    URLhttps://arxiv.org/abs/2407.03726

    doi: 10.48550/ARXIV.2407.03726. URLhttps://arxiv.org/abs/2407.03726

    doi arXiv

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.