REVIEW 4 major objections 4 minor 13 references
Integral of an alpha unpredictable function
T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash
Pith's one-line read A bounded integral of an alpha unpredictable function is again alpha unpredictable.
desk verdict The intended endpoint theorem is plausible, but the only proof given does not close; the flaws are repairable but should be fixed before anyone cites it. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central machinery is the method of included intervals, combined with sup/inf estimates on the primitive. To prove Poisson stability, the paper fixes a bounded interval $[a,b]$, chooses points $t_1,t_2$ at which $F$ is close to its supremum and infimum, encloses $[a,b]$ together with those points in a larger interval $[c,d]$, and uses uniform convergence of $f(t+t_n)$ on $[c,d]$ to force $F(t+t_n)-F(t)$ small on the larger interval and hence on $[a,b]$. For $\alpha$ unpredictability, the key identity is $F(t+t_n)-F(t)=F(s_n+t_n)-F(s_n)+\int_{s_n}^{t}(f(u+t_n)-f(u))\,du$; the two-case split locates a short interval, of length $h=\varepsilon_0\delta/(8M)$, on which the shifted difference stays above $\varepsilon_0\delta/4$.
What would settle it
Compute $F(t+t_n)-F(t)$ for a bounded continuous Poisson-stable integrand $f$ whose primitive $F$ approaches its supremum only as $t\to\infty$ and never reaches $M-\varepsilon/8$; if the difference fails to converge to zero uniformly on bounded intervals, Lemma 1 is false, and if it does converge, the attainment assumption is removable.
Extended reading notes
Core claim
The central claim is Theorem 1: if $f(t)$ is $\alpha$ unpredictable, then $F(t)=\int_{t_0}^{t}f(s)\,ds$ is $\alpha$ unpredictable if and only if $F$ is bounded. Lemma 1 establishes the Poisson-stability half: a bounded primitive of a Poisson stable function is Poisson stable, with the same convergence sequence $t_n$. The sufficiency proof for $\alpha$ unpredictability then splits into two cases depending on whether $|F(s_n+t_n)-F(s_n)|$ already exceeds $\varepsilon_0\delta/2$; in the first case the divergence interval for $F$ is $[s_n,s_n+h]$, and in the second it is $[s_n+\delta-h,s_n+\delta]$, where $h=\varepsilon_0\delta/(8M)$, and in both cases the divergence amplitude is $\varepsilon_0\delta/4$. Corollary 1 extends the result to primitives of the form $ct+g(t)$: if $g$ is bounded, then $g$ is $\alpha$ unpredictable.
Load-bearing premise
The proof of Lemma 1 assumes a bounded continuous function $F$ attains the particular values $M-\varepsilon/8$ and $m+\varepsilon/8$, where $M$ and $m$ are its highest and lowest values; bounded continuous functions need not attain their extremes, so these equalities can fail.
Editorial extensions
If this is right
- The chain of recurrence classes invariant under indefinite integration is now complete: periodic, quasi-periodic, almost periodic, Poisson stable, and alpha unpredictable functions all have the property that a bounded integral stays in the same class.
- If an alpha unpredictable function has a bounded primitive, the integral operator maps the class into itself; unbounded primitives are excluded.
- The paper's Corollary 1 gives a practical test: whenever an alpha unpredictable integrand has a primitive of the form $ct+g(t)$ with $g$ bounded, the bounded part $g$ is itself alpha unpredictable.
- The divergence data for the integral are explicit: given $\varepsilon_0,\delta$ for $f$, one can exhibit intervals of length $h=\varepsilon_0\delta/(8M)$ with divergence amplitude $\varepsilon_0\delta/4$.
Reading between the lines
- A repair of the attainment step in Lemma 1 via approximation would make the theorem cover bounded primitives whose supremum and infimum are not attained; the current proof text assumes the exact equalities $F(t_1)=M-\varepsilon/8$ and $F(t_2)=m+\varepsilon/8$.
- The two-case construction may iterate: higher-order primitives of alpha unpredictable functions should inherit alpha unpredictability as long as every intermediate primitive stays bounded, since each integration step can reuse the same divergence-sequence split.
- The explicit constants suggest a numerical test: for a concrete alpha unpredictable function one can measure the actual divergence amplitude and interval length of the integral and compare them with the paper's $\varepsilon_0\delta/4$ and $\varepsilon_0\delta/(8M)$.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims that the integral of an alpha unpredictable function is alpha unpredictable if and only if the integral is bounded. The main results are Lemma 1, which states that the integral of a Poisson stable function is Poisson stable exactly when it is bounded, and Theorem 1, which extends this to alpha unpredictability. The proofs use a so-called method of included intervals. The paper also derives a corollary for functions whose integral has a linear trend plus a bounded Part.
Significance. If the result is correct, it completes a natural hierarchy in the theory of recurrent functions: periodicity, quasi-periodicity, almost periodicity, Poisson stability, and alpha unpredictability. The boundedness condition is the known necessary and sufficient condition for the integral to preserve recurrence, so the statement is plausible and of interest to the community. The paper also advertises the method of included intervals as a verification tool. However, the current proofs contain several technical gaps that prevent the central claim from being established as written.
major comments (4)
- [Section 2, Lemma 1, equations (1)–(2)] The proof assumes that the continuous bounded function F attains values exactly equal to M−ε/8 and m+ε/8. This is not true in general; a continuous bounded function on R need not attain its supremum or infimum (e.g., arctan t). The subsequent estimates rely on these equalities, so the first step of the proof is invalid. This can be repaired by taking points with F(t1)>M−ε/8 and F(t2)<m+ε/8, but the constants in the rest of the proof must then be reworked.
- [Section 2, Lemma 1, epsilon estimates] The proof fixes an index n such that |f(t+t_n)−f(t)|<5ε/(8(d−c)) on [c,d], but later bounds the corresponding integral by ε(d−c)/(4(d−c))=ε/4. Using the stated 5ε/8 bound yields an error of 5ε/8, not ε/4. Consequently the final inequalities give |F(t+t_n)−F(t)|<11ε/8, not <ε. The epsilon chain does not close as written. The proof must either choose a stricter uniform closeness (for instance ε/(4(d−c))) or adjust all constants consistently.
- [Section 2, Theorem 1, case (ii)] The inequality |F(s_n+δ+t_n)−F(s_n+δ)|>ε0δ−ε0δ/2 relies on the implicit claim that |∫_{s_n}^{s_n+δ}(f(s+t_n)−f(s))ds|>ε0δ. This requires that f(s+t_n)−f(s) does not change sign on the interval, which follows from continuity and the condition |f(s+t_n)−f(s)|>ε0, but this continuity/sign argument is not stated. In addition, the proof uses h=ε0δ/(8M) without defining M in Theorem 1; M must be a uniform bound for |f|, and it should not be confused with the sup F from Lemma 1.
- [Section 2, Theorem 1, divergence sequence] The constructed divergence moments s'_n are chosen as either s_n or s_n+δ−h depending on the case. The resulting sequence may fail to be strictly increasing, as required by Definition 2; for example, if s'_n=s_n+δ−h and s'_{n+1}=s_{n+1}, the increments may not be positive unless s_{n+1}−s_n>δ−h. The proof should either thin to a subsequence where the choice is constant, or otherwise justify that the selected s'_n form a strictly increasing sequence tending to infinity.
minor comments (4)
- [Section 2, Theorem 1, proof] The proof of Theorem 1 contains a duplicated introductory paragraph: 'In what follows, we shall apply the characteristics of function f(t)...' appears twice almost verbatim.
- [Section 2, Lemma 1, necessity] The claim that necessity is an immediate consequence of Definition 1 is opaque, since Definition 1 does not state boundedness. The authors' opening sentence restricts attention to uniformly bounded functions, so the necessity direction is then trivial, but the link should be made explicit.
- [Abstract and Introduction] The phrase 'any theoretical functional recurrence is invariant with respect to integration' is broader than the theorem actually proved, which concerns alpha unpredictable functions. The abstract and introduction would benefit from a more precise statement of the scope.
- [References and typos] There are several typographical issues, such as 'effectivemethod', 'V .V ' in references, and inconsistent spacing in author names. The reference to the authors' prior work [12] is essential for the algebra used in Corollary 1; the paper should state the needed properties explicitly or reproduce them.
Circularity Check
No significant circularity: the main theorem is derived from the definition of alpha unpredictability and explicit epsilon-delta estimates; the only load-bearing self-citation is in Corollary 1, not in the central proof.
full rationale
The central derivation chain is not circular. Theorem 1 does not define alpha unpredictability in terms of the integral nor does it fit any parameter and call it a prediction. Lemma 1 is proved by an epsilon-delta argument using the sup and inf of F and the uniform convergence of shifts; Theorem 1 constructs explicit divergence intervals of length h = epsilon_0*delta/(8M) and amplitude epsilon_0*delta/4 directly from the alpha unpredictability of f. The citations to the authors' prior work [12] are used for the definition of alpha unpredictability, for the method of included intervals (which is then carried out in the proof), and for algebraic closure properties invoked in Corollary 1. None of these citations imports the main theorem's conclusion. The proof does contain genuine correctness gaps: Lemma 1 assumes F attains values equal to M - epsilon/8 and m + epsilon/8, which a continuous bounded function need not do, and the epsilon bookkeeping in the included-interval estimate is inconsistent. However, these are mathematical defects in the proof as written, not circularity. Corollary 1 leans on [12] for the claim that subtracting a constant preserves alpha unpredictability; that is a self-citation that is load-bearing only for the corollary, while the central theorem has independent content. Accordingly, the paper merits a low score, reflecting one minor self-citation rather than a circular derivation.
Assumptions & free parameters
assumptions (4)
- ad hoc to paper The integral F attains values exactly equal to M−ε/8 and m+ε/8, where M=sup F and m=inf F.
- domain assumption For t in [s_n, s_n+δ], |f(s+t_n)−f(s)| > ε0 implies |∫_{s_n}^{s_n+δ}(f(s+t_n)−f(s))ds| > ε0δ.
- ad hoc to paper The divergence sequence for F, obtained by choosing s_n or s_n+δ−h depending on the case, can be made strictly increasing.
- domain assumption Alpha unpredictable functions are closed under addition of constants and under subtraction (algebra from [12]).
Cite this review
Pith. "Pith review of Integral of an alpha unpredictable function." pith.science (2026). https://pith.science/paper/H3V3VOXY
@misc{pith2026260809809,
author = {Pith},
title = {Pith review of: Integral of an alpha unpredictable function},
year = {2026},
howpublished = {\url{https://pith.science/paper/H3V3VOXY}},
note = {Machine review of arXiv:2608.09809}
}
read the original abstract
The present paper is devoted to the study of integral properties of alpha unpredictable functions. The problem of invariance of recurrence under integration is one of the most challenging and interesting in the theory of functions. Let us start with periodicity. Next, the problem was solved for quasiperiodic, almost periodic, and Poisson stable functions. All of the problems were subjected to the condition of bounded integrals. The alpha-unpredictable functions are the endpoint in the row of recurrent functions. The class is the cross-border point from regularity to chaos in the dynamics presented through the elements of the row. In the present research, we finalized the proof that any theoretical functional recurrence is invariant with respect to integration, provided the result is bounded.
Reference graph
Works this paper leans on
-
[1]
Poincar ´e, H.; Sur le probl ´eme destrois corpsetles ´equations de la dynamique.Acta Math. 1890,13, 1–270
-
[2]
Poincar ´e, H.; Les methodes nouvelles de la mecanique celeste.Paris: Gauthier-Villars1892, V ol. 1, 2
-
[3]
Dynamical Systems.American Mathematical Society, Providence, RI, USA, 1927
Birkhoff, G. Dynamical Systems.American Mathematical Society, Providence, RI, USA, 1927
work page 1927
-
[4]
Bohl, P.; ¨Uber eine Differentialgleichung der St¨orungstheorie.Grelles J.,1900,131, 268–321
work page 1900
-
[5]
Esclangon, E.;Les Fonctions Quasi-Periodiques.Paris: Gauthier-Villars,1904
work page 1904
-
[6]
Bohr H.; Almost Periodic Functions.Chelsea Publishing Company, New York,1947
work page 1947
-
[7]
Levitan, B.M.; Zhikov, V .V .; Almost Periodic Functions and Differential Equations.Cam- bridge: Cambridge University Press,1982, pp. 328
work page 1982
-
[8]
Shcherbakov B.A.: Poisson stable solutions of differential equations, and topological dynam- ics.Differential Equations,5,1969, 2144–2155
work page 1969
Show all 13 references
-
[9]
Nemyckii, V .V and Stepanov, V .V .; Qualitative Theory of Differential Equations (Russian), Moscow, OGIZ,1947
1947
-
[10]
R.; Topological Dynamics and Ordinary Differential Equations.V an Nostrand Rein- hold Company, London,1971
Sell, G. R.; Topological Dynamics and Ordinary Differential Equations.V an Nostrand Rein- hold Company, London,1971
1971
-
[11]
and Fen M.O
Akhmet M. and Fen M.O. Unpredictable points and chaos.Journal Communication in Non- linear Science and Numerical Simulation,2016,40, 1–5
2016
-
[12]
Akhmet M.; Ultra Poincar ´e chaos and alpha labeling.IOP , Birmingham,2025
2025
-
[13]
U.,Baskan K
Akhmet M. U.,Baskan K. and Yesil C.: Delta synchronization of Poincar ´e chaos in gas discharge-semiconductor systems.Chaos,2022,32(81), 083137
2022
Reviewed August 11, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.