Using the so called monotonicity property, we prove that the Borel mapping restricted to some quasi-anlytic classes is never onto.
Paper ID : 1911.01302Title : On some quasi-analytic classesAuthor : Abdelhafed Elkhadiri (University Ibn Tofail, Faculty of Sciences, Kenitra, Morocco)Classification : math.FA (Functional Analysis)Publication Date : November 4, 2019Paper Link : https://arxiv.org/abs/1911.01302 This paper utilizes monotonicity properties to prove that the restriction of the Borel map on certain quasi-analytic classes is never surjective.
Analytic functions on an interval [ a , b ] ⊂ R [a,b] \subset \mathbb{R} [ a , b ] ⊂ R possess two equivalent important properties:
Property B (Bernstein definition) : An analytic function on [ a , b ] [a,b] [ a , b ] is completely determined by its values on subintervalsProperty DC (Denjoy-Carleman definition) : An analytic function on [ a , b ] [a,b] [ a , b ] is completely determined by its function values and all derivatives at a single point c ∈ [ a , b ] c \in [a,b] c ∈ [ a , b ] For a long time, it was believed that analytic functions were the only function class possessing property DC. It was not until Borel first proved the existence of more general function classes—quasi-analytic function classes—that also satisfy property DC.
The core problems addressed in this paper are:
Do quasi-analytic function classes preserve the monotonicity property ? Is the Borel map surjective on these classes? For an analytic function f f f , if all derivatives at a point c c c are non-negative: f ( n ) ( c ) ≥ 0 , ∀ n ∈ N f^{(n)}(c) \geq 0, \forall n \in \mathbb{N} f ( n ) ( c ) ≥ 0 , ∀ n ∈ N , then there exists η > 0 \eta > 0 η > 0 such that f ( n ) ( x ) ≥ 0 , ∀ n ∈ N , ∀ x ∈ [ c , η ] f^{(n)}(x) \geq 0, \forall n \in \mathbb{N}, \forall x \in [c, \eta] f ( n ) ( x ) ≥ 0 , ∀ n ∈ N , ∀ x ∈ [ c , η ] .
Proved the validity of the monotonicity property for Denjoy-Carleman quasi-analytic classes Provided a new proof of Carleman's theorem using the monotonicity property : proving that the Borel map restriction on quasi-analytic classes is never surjectiveExtended Bang's theorem to quasi-analytic classes related to integer sequencesProvided direct proof methods that do not rely on functional analysis techniquesLet C ∞ ( [ a , b ] ) C^{\infty}([a,b]) C ∞ ([ a , b ]) denote the ring of infinitely differentiable functions on the interval [ a , b ] [a,b] [ a , b ] , and c ∈ [ a , b ] c \in [a,b] c ∈ [ a , b ] . The Borel map is defined as:
T c : C ∞ ( [ a , b ] ) → R [ [ x ] ] T_c: C^{\infty}([a,b]) \to \mathbb{R}[[x]] T c : C ∞ ([ a , b ]) → R [[ x ]] f ↦ ∑ n = 0 ∞ f ( n ) ( c ) n ! x n f \mapsto \sum_{n=0}^{\infty} \frac{f^{(n)}(c)}{n!} x^n f ↦ ∑ n = 0 ∞ n ! f ( n ) ( c ) x n
The research objective is to prove that the restriction of this map on quasi-analytic classes is not surjective.
Given a sequence of positive numbers M = ( M n ) n ∈ N M = (M_n)_{n \in \mathbb{N}} M = ( M n ) n ∈ N , the Denjoy-Carleman class C M ( [ a , b ] ) C_M([a,b]) C M ([ a , b ]) is defined as:
C M ( [ a , b ] ) = { f ∈ C ∞ ( [ a , b ] ) : ∣ f ( n ) ( x ) ∣ ≤ c ⋅ C n M n , ∀ n ∈ N , ∀ x ∈ [ a , b ] } C_M([a,b]) = \{f \in C^{\infty}([a,b]): |f^{(n)}(x)| \leq c \cdot C^n M_n, \forall n \in \mathbb{N}, \forall x \in [a,b]\} C M ([ a , b ]) = { f ∈ C ∞ ([ a , b ]) : ∣ f ( n ) ( x ) ∣ ≤ c ⋅ C n M n , ∀ n ∈ N , ∀ x ∈ [ a , b ]}
where c , C c, C c , C are positive constants depending on f f f but independent of n n n .
Theorem 2.6 : The class C M ( [ a , b ] ) C_M([a,b]) C M ([ a , b ]) is quasi-analytic if and only if ∑ n = 0 ∞ 1 β n = ∞ \sum_{n=0}^{\infty} \frac{1}{\beta_n} = \infty ∑ n = 0 ∞ β n 1 = ∞ , where β n = inf k ≥ n M k k \beta_n = \inf_{k \geq n} \sqrt[k]{M_k} β n = inf k ≥ n k M k .
To handle general sequences, logarithmically convex regularization is introduced:
Definition 2.7 : A sequence M = ( M n ) n ∈ N M = (M_n)_{n \in \mathbb{N}} M = ( M n ) n ∈ N is called logarithmically convex if for all n ≥ 1 n \geq 1 n ≥ 1 , M n 2 ≤ M n − 1 M n + 1 M_n^2 \leq M_{n-1} M_{n+1} M n 2 ≤ M n − 1 M n + 1 .
Through Newton polygon construction, any sequence satisfying lim inf n → ∞ M n n = ∞ \liminf_{n \to \infty} \sqrt[n]{M_n} = \infty lim inf n → ∞ n M n = ∞ admits a logarithmically convex regularization M c = ( M n c ) n ∈ N M^c = (M_n^c)_{n \in \mathbb{N}} M c = ( M n c ) n ∈ N .
Theorem 2.10 : The following conditions are equivalent:
∑ n = 0 ∞ 1 β n = ∞ \sum_{n=0}^{\infty} \frac{1}{\beta_n} = \infty ∑ n = 0 ∞ β n 1 = ∞ ∑ n = 0 ∞ 1 M n c n = ∞ \sum_{n=0}^{\infty} \frac{1}{\sqrt[n]{M_n^c}} = \infty ∑ n = 0 ∞ n M n c 1 = ∞ ∑ n = 0 ∞ M n − 1 c M n c = ∞ \sum_{n=0}^{\infty} \frac{M_{n-1}^c}{M_n^c} = \infty ∑ n = 0 ∞ M n c M n − 1 c = ∞ Construction of a metric on the real sequence space S ( R ) S(\mathbb{R}) S ( R ) :
∥ X ∥ = inf k ∈ P ( max ( e − k , max 0 ≤ n ≤ k ∣ x n ∣ ) ) \|X\| = \inf_{k \in P} \left(\max\left(e^{-k}, \max_{0 \leq n \leq k} |x_n|\right)\right) ∥ X ∥ = inf k ∈ P ( max ( e − k , max 0 ≤ n ≤ k ∣ x n ∣ ) )
Lemma 3.6 : If ∥ X f ( t ) ∥ ≠ 0 \|X_f(t)\| \neq 0 ∥ X f ( t ) ∥ = 0 , then
∥ X f ( t + τ ) ∥ ≤ ∥ X f ( t ) ∥ exp ( e ∣ τ ∣ M l c M l − 1 c ) \|X_f(t + \tau)\| \leq \|X_f(t)\| \exp\left(e|\tau| \frac{M_l^c}{M_{l-1}^c}\right) ∥ X f ( t + τ ) ∥ ≤ ∥ X f ( t ) ∥ exp ( e ∣ τ ∣ M l − 1 c M l c )
Theorem 4.2 : Let f ∈ C ∞ ( [ a , b ] ) f \in C^{\infty}([a,b]) f ∈ C ∞ ([ a , b ]) satisfy sup t ∈ [ a , b ] ∣ f ( n ) ( t ) ∣ ≤ M n \sup_{t \in [a,b]} |f^{(n)}(t)| \leq M_n sup t ∈ [ a , b ] ∣ f ( n ) ( t ) ∣ ≤ M n . If there exists a sequence ( x n ) n (x_n)_n ( x n ) n such that f ( n ) ( x n ) = 0 , ∀ n ∈ N f^{(n)}(x_n) = 0, \forall n \in \mathbb{N} f ( n ) ( x n ) = 0 , ∀ n ∈ N , and the series ∑ n = 1 ∞ ∣ x n − 1 − x n ∣ \sum_{n=1}^{\infty} |x_{n-1} - x_n| ∑ n = 1 ∞ ∣ x n − 1 − x n ∣ converges, then f ≡ 0 f \equiv 0 f ≡ 0 .
This paper is primarily theoretical research, with conclusions verified through rigorous mathematical proofs. Main verification contents include:
Proof of sufficiency in Carleman's theorem (Section 3)Verification of monotonicity property (Section 4)Properties of quasi-analytic classes related to integer sequences (Sections 5-6)Use proof by contradiction: assume the existence of a non-zero function satisfying the conditions Construct decreasing sequences and apply the intermediate value theorem Utilize series divergence to derive a contradiction Theorem 4.7 : Let C M ( [ a , b ] ) C_M([a,b]) C M ([ a , b ]) be a quasi-analytic class that strictly contains the analytic class. Then the Borel map T c : C M ( [ a , b ] ) → R [ [ x ] ] T_c: C_M([a,b]) \to \mathbb{R}[[x]] T c : C M ([ a , b ]) → R [[ x ]] is not surjective.
Proof Strategy : Consider a non-convergent series ∑ n ∈ N a n x n \sum_{n \in \mathbb{N}} a_n x^n ∑ n ∈ N a n x n (where a n > 0 a_n > 0 a n > 0 ). Using the monotonicity property, one can prove that this series is not in the image of T c ( C M ( [ a , b ] ) ) T_c(C_M([a,b])) T c ( C M ([ a , b ])) .
Corollary 4.5 : Let f ∈ C ∞ ( [ a , b ] ) f \in C^{\infty}([a,b]) f ∈ C ∞ ([ a , b ]) satisfy sup t ∈ [ a , b ] ∣ f ( n ) ( t ) ∣ ≤ M n \sup_{t \in [a,b]} |f^{(n)}(t)| \leq M_n sup t ∈ [ a , b ] ∣ f ( n ) ( t ) ∣ ≤ M n . If f ( n ) ( a ) > 0 f^{(n)}(a) > 0 f ( n ) ( a ) > 0 for all n ∈ N n \in \mathbb{N} n ∈ N , then f ( n ) ( x ) > 0 f^{(n)}(x) > 0 f ( n ) ( x ) > 0 for all n ∈ N n \in \mathbb{N} n ∈ N and x ∈ [ a , b ] x \in [a,b] x ∈ [ a , b ] .
Theorem 6.2 : For quasi-analytic classes related to integer sequences n = ( n k ) k ∈ N n = (n_k)_{k \in \mathbb{N}} n = ( n k ) k ∈ N , if f ( n ) ( x n ) = 0 f^{(n)}(x_n) = 0 f ( n ) ( x n ) = 0 and ∑ j = 0 ∞ ∣ x j − x j + 1 ∣ \sum_{j=0}^{\infty} |x_j - x_{j+1}| ∑ j = 0 ∞ ∣ x j − x j + 1 ∣ converges, then f ≡ 0 f \equiv 0 f ≡ 0 .
Universality of the monotonicity property : It holds not only for analytic functions but also for broader quasi-analytic classesLimitations of the Borel map : It can never achieve surjectivity on quasi-analytic classesSpecial status of geometric series : Divergent power series with positive coefficients cannot be realized through quasi-analytic functionsBorel (1917) : First introduced the concept of quasi-analytic functionsBernstein (1914) : Adopted property B to define quasi-analyticityDenjoy-Carleman : Adopted property DC definition and provided complete theoryCarleman (1926) : Proved that the Borel map is non-surjective on quasi-analytic classesBang (1953) : Applied metric space theory to study quasi-analytic classesCompared to existing work, this paper:
Provides direct proofs that do not rely on functional analysis Extends the monotonicity property to broader quasi-analytic classes Establishes new quasi-analytic class theory related to integer sequences Preservation of monotonicity property : Denjoy-Carleman quasi-analytic classes preserve the monotonicity property of analytic functionsNon-surjectivity of the Borel map : Carleman's theorem is proved using the monotonicity propertyExtensibility of the theory : The method can be generalized to quasi-analytic classes related to integer sequencesRestricted scope of applicability : Primarily applicable to function classes satisfying specific growth conditionsLack of constructivity : The proofs are non-constructive and do not provide explicit constructions of elements outside the imageComputational complexity : Practical verification of quasi-analyticity conditions may be computationally complexThe paper raises an open problem:
Problem 6.5 : Do all quasi-analytic classes satisfy the monotonicity property? In particular, does the monotonicity property hold for definable functions in polynomial-bounded o-minimal structures?
Theoretical depth : Provides new perspectives and deep understanding of quasi-analytic function theoryProof techniques : Cleverly employs metric space theory and monotonicity propertiesCompleteness of results : Not only proves the main theorems but also provides multiple corollaries and extensionsHistorical continuity : Well connects with classical work by Carleman and othersLimited applicability : Primarily theoretical results with unclear practical applicationsComputational aspects : Lacks effective algorithms to determine whether specific functions belong to certain quasi-analytic classesInsufficient examples : Lacks concrete numerical examples to illustrate theoretical resultsTheoretical contribution : Provides new research tools and perspectives for quasi-analytic function theoryMethodological innovation : Application of monotonicity properties offers new approaches to related problemsFuture research directions : Provides new directions for research in o-minimal structures and related modern mathematicsFunctional analysis theory research : Provides tools for studying analytic continuation properties of functionsDifferential equation theory : May have applications in studying analyticity properties of solutionsComplex analysis : Provides theoretical foundation for studying analytic continuation of real analytic functions to the complex planeThe paper cites important literature in the field, including:
Carleman's classical work Les fonctions quasi-analytiques (1926) Mandelbrojt's Séries adhérentes (1952) Bang's application of metric space theory (1953) Modern o-minimal structure theory literature These references demonstrate the author's deep understanding of the historical development and current state of the field.