{"id":"0531ed28-568f-44a5-ab1f-9ead7703134b","arxiv_id":"1908.05115","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A new Schur-type algorithm for matricial Hausdorff moment sequences on [alpha,beta] reduces length by one and acts as a left shift on the associated sequence of canonical moments.","lead":"This mathematics paper builds a new transformation, the F-alpha,beta-transform, that shortens a Hausdorff moment sequence on an interval by one step while preserving its defining positivity property. It then shows this transform acts as a left shift on the sequence's canonical moments, and uses it to characterize balanced (central) matrix-valued measures via the arcsine distribution.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Schur-algorithm interpretation of Theorem 9.14 rests on the imported bijection [24, Thm. 6.30]; no internal flaw found, but that dependency is the main risk.","rationale":"The reader correctly identified the imported bijection theorem as the weakest assumption. I read the proof of Theorem 9.14 in full and found no internal inconsistency: the interval-parameter sequence of the transformed sequence is defined constructively, and the equality p_j = e_{k+j} follows from the F-parameter shift in Theorem 9.13 together with the interval-length scaling in Proposition 9.11. The bijection is not needed for that computation. It is, however, needed for the claim that the F-transform is a genuine Schur-type algorithm in the full matricial Hausdorff moment space and for the arcsine characterization in Section 10. Since the bijection is imported from a published paper and no counterexample or internal reason to doubt it appears, I do not see grounds to lower the verdict; the appropriate status remains ACCEPT with moderate confidence. A targeted degenerate-case verification of the bijection would be the single most useful check.","tokens_in":81938,"tokens_out":22021,"duration_ms":215529,"concrete_test":"Test Theorem 7.34 in the degenerate q=2 case: take d0 = diag(1,0), choose an admissible interval-parameter sequence e in E^nonneg_{2,kappa,delta} with e_j having the correct projection (e.g. e1 = lambda diag(1,0)), reconstruct the candidate moment sequence via the explicit formulas in [24, Def. 6.21/Notation 6.28], and check all finite Hankel matrices H_n, H_{alpha,n,dot}, H_{dot,n,beta}, H_{alpha,n,beta} are nonnegative Hermitian and the reconstruction is unique. If reconstruction fails or is non-unique, Theorem 9.14's Schur-algorithm interpretation is undermined; success for several random degenerate parameter sequences would settle the concern.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 9.14 asserts that the k-th F_{alpha,beta}-transform has interval-parameter sequence p0 = delta^{k-1} d_k and p_j = e_{k+j}. The sequence p is well-defined directly from Definition 7.32 and the proof goes through Theorem 9.13 and Proposition 9.11, so the algebraic equality itself does not invoke Theorem 7.34. The load-bearing dependency is the advertised meaning: interval parameters are the canonical-moment coordinates only because of the bijection Sigma_{alpha,beta} in Theorem 7.34 (imported from [24, Thm. 6.30]), and the measure-level counterpart in Theorem 10.3. This bijection includes degenerate, rank-deficient Hankel matrices; if it had a hidden counterexample there, the left-shift statement would not determine a unique moment sequence from the shifted parameters, and the proof of Theorem 10.9 would fail where it invokes Theorem 10.3. The paper does not re-prove the bijection, and no independent verification is supplied.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an algebraic Schur-type algorithm for the matricial Hausdorff moment problem on a compact interval [α, β]. It constructs the F_{α,β}-transform of a finite or infinite sequence of complex matrices and proves, in Theorem 9.4, that the transform preserves the class of [α, β]-non-negative definite sequences, reducing the length by one. The main structural result, Theorem 9.14, identifies the interval-parameter sequence of the k-th F_{α,β}-transform: its initial entry is δ^{k-1} d_k and its remaining entries are the original interval parameters shifted by k. This is interpreted as an essentially left shift of the matricial canonical-moment parameters. In Section 10 the authors apply the transform to non-negative Hermitian measures on [α, β] and characterize molecular measures and centrality: a measure is central of order k precisely when its (k-1)-st transform has the matricial arcsine form stated in Theorem 10.9. The technical core consists of long block-Hankel and Schur-complement computations in Sections 4, 8, and 9, including explicit rank and determinant formulas.","tokens_in":82092,"tokens_out":5801,"duration_ms":65109,"significance":"If the results are correct, this is a substantial and novel contribution: it provides the first Schur-type algorithm for the matricial Hausdorff moment problem on a compact interval, generalizing the classical scalar canonical-moment theory to the degenerate, non-invertible Hankel case. The proof of the left-shift identity is, conditional on previously published parametrization results, self-contained and detailed, and the paper also supplies explicit block-LDU factorizations and rank/determinant identities that are likely to be useful for further work. The applications to molecular measures and to the arcsine-distribution characterization of centrality are natural and give the algebraic machinery a concrete measure-theoretic payoff. The main caveat, which the authors themselves make transparent, is the dependence on the bijection in Theorem 7.34 (imported from [24, Thm. 6.30]) for the interpretation of interval parameters as canonical moments; this is a published result with a proof, so the dependence is acceptable, but it should be flagged explicitly in the statement of Theorem 9.14.","major_comments":[],"minor_comments":[{"comment":"The interpretation of the interval-parameter sequence as a coordinate system for the moment space relies on the bijection in Theorem 7.34 (imported from [24, Thm. 6.30]). The algebraic computation of p_j in the proof of Theorem 9.14 is self-contained, but the statement and the paragraph following it should explicitly say that uniqueness of the shifted parameter sequence is part of the imported bijection; this is a clarification, not a gap.","section":"Theorem 9.14 and surrounding discussion"},{"comment":"The same letter F is used for the F_{α,β}-parameter sequence and for the F_{α,β}-transform. Although the context usually disambiguates them, the notation is heavy throughout the paper and this collision makes Sections 8 and 9 harder to read; consider renaming one of the two objects.","section":"Definitions 7.27 and 8.14"},{"comment":"The formula p_0 = δ^{k-1} d_k is stated for all k ∈ Z_{0,κ}; for k = 0 this involves δ^{-1}, which is well defined because δ > 0, but the k = 0 case should be stated separately or accompanied by a convention for negative powers of δ.","section":"Theorem 9.14 and Proposition 10.4"},{"comment":"The proof is very long and uses equations (8.36), (8.41), and (8.44) in a somewhat intricate way; adding a short sentence after (8.25) explaining the role of these three identities in the reduction to the rank and determinant formulas would improve readability.","section":"Proof of Proposition 8.38"},{"comment":"There are minor typographical inconsistencies in the rendering of the block-Toeplitz inverse notation, with S♯ and S† both appearing in closely related roles in Propositions 8.38 and 8.40; harmonizing this notation would prevent confusion for readers who track the Moore-Penrose inverse through the proofs.","section":"Sections 5 and 8"}],"recommendation":"minor_revision","confidential_remarks":"This is a competent continuation of the authors' long-term program on matricial moment problems. The main theorems are supported by extensive calculations, and the reliance on the previously published bijection [24, Thm. 6.30] is legitimate and correctly cited. The only caveat is that the meaning of the central left-shift theorem is fully dependent on that imported parametrization; if the editors are aware of any unresolved issue with degenerate cases in [24], that would be worth double-checking, but nothing in the present manuscript indicates such a problem. The paper is within the scope of the journal and the presentation, while notationally demanding, is consistent with the authors' previous work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth knowing first: this is the compact-interval instance of the authors' Schur-type algorithm for matricial moment sequences, and it is a genuine new result, not a repackaging. The main theorem (9.14) says the k-th F_{alpha,beta}-transform of an [alpha,beta]-non-negative definite sequence is again one, and its [alpha,beta]-interval parameters are p_0 = delta^{k-1} d_k and p_j = e_{k+j}. That is exactly a left shift of the canonical-moment coordinates, so the Schur-reading is justified. I checked the line of proof and found no internal contradiction; the extensive block Hankel and LDU identities are worked out in detail and look consistent.\n\nThe genuinely new piece is the F_{alpha,beta}-transform (Definition 8.14) and its iterated use. Previous papers in the series treated R and half-lines; the compact interval had no such algorithm, and I find the claim that even the scalar case lacked it plausible. The treatment of degenerate, non-invertible Hankel blocks via Moore-Penrose inverses is a real plus over older scalar canonical-moment treatments. The final application, Theorem 10.9, tying centrality to a matricial arcsine distribution after k steps, is a clean payoff.\n\nThe soft spot is not in the algebra I could see; it is a dependency. Theorem 9.14's advertised meaning as a Schur algorithm inherits the bijection theorem [24, Thm. 6.30] (Theorem 7.34 here), which includes the degenerate case. The paper does not re-prove that bijection. If that theorem had a hidden counterexample, the shifted parameter sequence would not uniquely determine a moment sequence, and Theorem 10.9 would fail where it invokes Theorem 10.3. That is a reasonable caution, not a demonstrated gap. The algebraic equality p_j = e_{k+j} itself is derived from local definitions and does not need the bijection; only the interpretation does. The stress-test concern is therefore real but not fatal.\n\nThe citation pattern is heavily self-referential, but the cited results are published with proofs, so I do not count that against it. The paper is long—over sixty pages—and human checking is the only verification; no machine-checked proofs or code. That caps confidence but does not undermine the result as far as I read.\n\nWho it is for: specialists in matricial moment problems, canonical moments, and Schur-type algorithms. It deserves a serious referee. I would send it to peer review and let referees spend the time; the main thing to verify is the imported bijection and the lengthy identities in Section 8.","headline":"A genuine, carefully worked Schur-type algorithm for matricial Hausdorff moment sequences on a compact interval; the main new theorem appears correct, with the main risk being the imported—not reproved—parametrization bijection from earlier work.","tokens_in":82689,"tokens_out":3343,"would_cite":true,"duration_ms":37206,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["44A60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper develops a Schur-type algorithm for matricial Hausdorff moment problems and proves that its core transform is a left shift of canonical-moment parameters.","keywords":["matricial Hausdorff moment problem","Schur–Nevanlinna type algorithm","matricial canonical moments","non-negative Hermitian block Hankel matrices","moments of matrix-valued measures","arcsine distribution","F_alpha,beta-transform","matrix measures on a compact interval"],"falsifier":"Compute the $F_{0,1}$-transform of a scalar ($q=1$) three-point atomic probability measure on $[0,1]$, list its moment sequence and interval parameters $(e_j)$, and check that the transformed sequence is again a moment sequence whose interval parameters are $p_0 = d_1$ and $p_j = e_{1+j}$; a single mismatch in this degenerate case, where some $d_k$ is the zero matrix, would falsify Theorem 9.14.","tokens_in":81694,"feed_emoji":"📊","tokens_out":10625,"duration_ms":92754,"temperature":0.7,"pith_summary":"Matrix measures on an interval are encoded by their moment sequences; this paper develops the algebraic engine for a Schur-type algorithm on such sequences. The main object is a transformation, the $F_{\\alpha,\\beta}$-transform, that shortens any $[\\alpha,\\beta]$-non-negative definite sequence by one term while keeping it a moment sequence of a matrix measure on $[\\alpha,\\beta]$. Iterating it peels off one matrix parameter at a time, and the central theorem says this peeling is exactly a left shift of the matricial canonical-moment parameters: after $k$ steps the new parameters begin with $\\delta^{k-1}d_k$ and then continue with the original $e_{k+1}, e_{k+2}, \\dots$, where $\\delta = \\beta-\\alpha$. The payoff is structural: finite-support measures are exactly those of which some transform is the zero measure, and a measure is central precisely when one of its transforms is a matricial arcsine distribution.","feed_headline":"Hausdorff moment sequences shift left one step at a time","feed_subtitle":"A transform that preserves matrix moment sequences ties centrality to the arcsine law.","key_machinery":"The load-bearing object is the $F_{\\alpha,\\beta}$-transform of a finite or infinite sequence of complex matrices, defined using the reciprocal sequence (a recursive construction with Moore–Penrose inverses) and the Cauchy product. Around it the paper organizes four block Hankel matrices $H_n$, $H_{\\alpha,n,\\bullet}$, $H_{\\bullet,n,\\beta}$, and $H_{\\alpha,n,\\beta}$ built from the sequence and from its modifications for $[\\alpha,\\infty)$, $(-\\infty,\\beta]$, and $[\\alpha,\\beta]$. The algebraic core consists of identities (Propositions 8.38–8.45) that express the Hankel matrices of the transformed sequence as Schur-complement factorizations of the input's Hankel matrices; these identities prove that non-negative definiteness survives the transform. The interval parameters $(e_j)$, recursively defined from the interval lengths, are the matricial generalization of the classical canonical moments, and Theorem 9.14 identifies the transform with their left shift.","core_discovery":"The paper's central claim is Theorem 9.14: if $(s_j)$ is $[\\alpha,\\beta]$-non-negative definite with interval-parameter sequence $(e_j)$ and interval lengths $(d_j)$, then every iterated $F_{\\alpha,\\beta}$-transform is again $[\\alpha,\\beta]$-non-negative definite, and its interval-parameter sequence $(p_j)$ satisfies $p_0 = \\delta^{k-1}d_k$ and $p_j = e_{k+j}$ for $j \\ge 1$. Thus the transform shifts the canonical-moment data left, discarding the information already consumed and rescaling the new leading parameter by the interval length $\\delta = \\beta-\\alpha$. This is the Hausdorff-interval analogue of the classical Schur algorithm for functions on the unit disk and of the Nevanlinna treatment of the Hamburger moment problem. For measures, the same statement says that after $k$ transforms the remaining measure captures exactly the tail of the original canonical moments. Section 10 applies this to two geometric properties: a measure is molecular (finite support) exactly when some transform is the zero matrix measure, and central of order $k$ exactly when the $(k-1)$-th transform is, up to a scale factor, the arcsine density on $[\\alpha,\\beta]$ multiplied by a non-negative Hermitian matrix.","pith_inferences":["The $\\delta$-scaling in the transform suggests that a suitably rescaled iterated transform should converge to a stationary object for generic input measures; the paper does not explore this probabilistic reading.","Because the transform uses only Moore–Penrose inverses and finite matrix arithmetic, it is a candidate for a numerically implementable recursion that computes matricial canonical moments from empirical moments; this remains to be tested.","The centrality result hints at a family of distinguished matrix measures: replacing the half-projection $\\tfrac12 P_{R(d)}$ by other convex combinations of $0$ and $P_{R(d)}$ should yield explicit non-arcsine 'balanced' measures with the same left-shift behaviour."],"forward_implications":["Iterating the $F_{\\alpha,\\beta}$-transform yields block LDU factorizations of all four associated Hankel matrices, with the diagonal blocks read directly from the transforms (Lemmas 9.6 and 9.7).","The interval-length sequence of the $k$-th transform is $\\delta^k d_{k+j}$, so a sequence completely degenerate at order $\\ell$ becomes completely degenerate at order $\\max\\{0,\\ell-k\\}$ (Proposition 9.11 and Corollary 9.12).","For matrix measures, molecularity is characterized by some $M[\\alpha,\\beta]$-transform being the zero measure (Proposition 10.5).","Centrality of order $k$ is characterized by the $(k-1)$-th transform being a scalar arcsine-type density times $\\delta^{k-2}d_{k-1}$ (Theorem 10.9).","Since the interval-parameter map is a bijection, the transform parametrizes the matricial Hausdorff moment space step by step, one matrix parameter at a time."],"supporting_citations":[{"why":"Supplies the bijection between $[\\alpha,\\beta]$-non-negative definite sequences and interval-parameter sequences, the parametrization in which the main theorem is stated.","marker":"[24]"},{"why":"Establishes the structural facts about interval lengths, Schur complements, and one-step extensions used to prove preservation of non-negative definiteness.","marker":"[23]"},{"why":"Provides the Hamburger-space template for the Schur-type transform and the canonical Hankel parametrization whose block-matrix machinery is adapted.","marker":"[26]"},{"why":"Introduces the matricial canonical-moment parametrization for intervals that the interval parameters generalize.","marker":"[13]"},{"why":"Gives the scalar canonical-moment theory and the arcsine-distribution example extended in the centrality theorem.","marker":"[12]"},{"why":"Develops reciprocal sequences and matrix-sequence invertibility via Moore–Penrose inverses, the technical foundation of the transform definition.","marker":"[27]"}],"fun_headline_variants":["Schur transform shifts matrix Hausdorff moments left","Hausdorff sequences shorten via canonical moment shift","Arcsine centrality via Schur analysis of moments","Left shift of canonical moments preserves Hausdorff type"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes as given the bijection theorem (Theorem 7.34) that every $[\\alpha,\\beta]$-non-negative definite sequence has a uniquely determined interval-parameter sequence, including degenerate cases in which the Hankel blocks are not invertible; if that parametrization failed, the left-shift statement in Theorem 9.14 would not have a well-defined conclusion.","fun_headline_variants_meta":{"raw":{"variants":["Schur transform shifts matrix Hausdorff moments left","Hausdorff sequences shorten via canonical moment shift","Arcsine centrality via Schur analysis of moments","Left shift of canonical moments preserves Hausdorff type"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000194,"raw_usage":{"total_tokens":1352,"prompt_tokens":942,"completion_tokens":410,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":558,"completion_tokens_details":{"reasoning_tokens":347}},"tokens_in":558,"tokens_out":410,"duration_ms":4469,"temperature":1.0,"reasoning_tokens":347,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:22:39.400508+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the $F_{0,1}$-transform of a scalar ($q=1$) three-point atomic probability measure on $[0,1]$, list its moment sequence and interval parameters $(e_j)$, and check that the transformed sequence is again a moment sequence whose interval parameters are $p_0 = d_1$ and $p_j = e_{1+j}$; a single mismatch in this degenerate case, where some $d_k$ is the zero matrix, would falsify Theorem 9.14.","supporting_citations":[{"cited_title":"Kirstein, and C","cited_arxiv_id":null,"evidence_quote":"Supplies the bijection between $[\\alpha,\\beta]$-non-negative definite sequences and interval-parameter sequences, the parametrization in which the main theorem is stated."},{"cited_title":"Kirstein, and C","cited_arxiv_id":null,"evidence_quote":"Establishes the structural facts about interval lengths, Schur complements, and one-step extensions used to prove preservation of non-negative definiteness."},{"cited_title":"Kirstein, C","cited_arxiv_id":null,"evidence_quote":"Provides the Hamburger-space template for the Schur-type transform and the canonical Hankel parametrization whose block-matrix machinery is adapted."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the matricial canonical-moment parametrization for intervals that the interval parameters generalize."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the scalar canonical-moment theory and the arcsine-distribution example extended in the centrality theorem."},{"cited_title":"Kirstein, C","cited_arxiv_id":null,"evidence_quote":"Develops reciprocal sequences and matrix-sequence invertibility via Moore–Penrose inverses, the technical foundation of the transform definition."}],"review_version":1}