Professional Documents
Culture Documents
A Sequence of Appell Polynomials and The Associate
A Sequence of Appell Polynomials and The Associate
https://doi.org/10.1007/s13324-021-00525-0
Abstract
We are concerned with a special sequence of Appell polynomials, related to the Rényi
and Tsallis entropies for the binomial distribution. The generating function is investi-
gated: it is logarithmically convex and has remarkable connections with the modified
Bessel function I0 (t) and with the index of coincidence for Poisson distribution. The
specific form of the Appell polynomials leads to specific properties of the associated
Jakimovski–Leviatan operators.
1 Introduction
B Ana-Maria Acu
anamaria.acu@ulbsibiu.ro
Ioan Cristian Buscu
buscu@yahoo.com
Ioan Rasa
ioan.rasa@math.utcluj.ro
1 Department of Mathematics and Informatics, Lucian Blaga University of Sibiu, Str. Dr. I. Ratiu,
No.5-7, 550012 Sibiu, Romania
2 Department of Mathematics, Faculty of Automation and Computer Science, Technical University of
Cluj-Napoca, Str. Memorandumului nr. 28, 400114 Cluj-Napoca, Romania
x
2n
Un (x) = n Un−1 (t)dt + , n ≥ 1, (1.1)
0 n
which was considered in [18] in relation with the sum of the squared fundamental Bern-
stein polynomials, and with Rényi and Tsallis entropies associated with the binomial
distribution.
From (1.1) we get by induction
n
n 2k
Un (x) = x n−k , n ≥ 0. (1.2)
k k
k=0
so that (Un (x))n≥0 is a sequence of Appell polynomials (see [7] and the references
therein).
In this paper we investigate the properties of the sequence (Un (x))n≥0 ; in particular,
its generating function and the Jakimovski–Leviatan type operators associated with
it. In Sect. 2 we consider the generating function a(t) corresponding to the sequence
of Appell polynomials (Un (x))n≥0 . Its explicit expression is given in (2.6). Using it
we derive a differential equation satisfied by a(t), which will be useful in the sequel.
The Laplace transform of a(t) is presented in Theorem 2.2, for which we give two
proofs. Using the Laplace transform we obtain the integral representation (2.11) of
a(t), as well as the relation between a(t) and the modified Bessel function I0 (t).
It is worth pointing out that (as the reviewers remarked) one can go in the opposite
direction: starting with the relation between a(t) and I0 (t), already known in a different
setting (see Remark 2.1), and using the properties of I0 (t) it is possible to obtain the
integral representation of a(t), the differential equation satisfied by it, and its Laplace
transform. The section ends with the relation between the generating function a(t)
and the index of coincidence for Poisson distribution. Using it we show that a(t) is
logarithmically convex on [0, ∞) and, moreover, we provide bounds for it.
Since 1969, when Jakimovski and Leviatan [14] published the paper, the study
of Jakimovski–Leviatan operators became a distinct topic in Approximation Theory.
These operators generalize the classical Favard–Szász–Mirakyan operators and, in
turn, were generalized in many directions. Their construction depends on the choice
of a sequence of Appell polynomials and one may take this sequence according to
the desired properties of the operators. Details can be found in [1,11,15,19,20] and
the references therein. Section 3 is devoted to the Jakimovski–Leviatan type operators
associated with the Appell polynomials Un (x). It is well known that if we consider
the operators associated with an arbitrary sequence of Appell polynomials, then the
corresponding moments depend on the successive derivatives of generating function
evaluated at 1. Theorem 3.1 shows that in our case the moments of the Jakimovski–
a (1)
Leviatan operators depend only on . Theorem 3.2 and Corollary 3.1 provide
a(1)
recurrence formulas for computing these moments. An application of Korovkin’s the-
ory in order to study the convergence of the operators to the identity operator ends this
section.
Concluding remarks and projects for further work are presented in Sect. 4.
According to the general theory (see, e.g., [7]), the generating function of the sequence
of Appell polynomials (An (x))n≥0 given by the explicit representation
n n
An (x) = αn + αn−1 x + αn−2 x 2 + · · · + α0 x n , n = 0, 1, . . . , (2.1)
1 2
t t2 tn
a(t) = α0 + α1 + α2 + · · · + αn + · · · , α0 = 0. (2.2)
1! 2! n!
t t2 tn
a(t)et x = A0 (x) + A1 (x) + A2 (x) + · · · + An (x) + · · · (2.3)
1! 2! n!
From (2.1) and (2.4) we see that for the sequence (Un (x))n≥0 the coefficients αi are
2k
αk = , k = 0, 1, 2, . . . (2.5)
k
The radius of convergence of this power series is infinite, and consequently a(t) is
analytic on R.
In what follows we will present some properties of the generating function a(t).
Theorem 2.1 The generating function a(t) satisfies the differential equation
we get
∞
∞
1 2k + 2 k 1 2k + 1 2k k
a (t) = t =2 t .
k! k + 1 k! k + 1 k
k=0 k=0
This implies
∞
1 2k k
a (t) = 4a(t) − 2 t ,
(k + 1)! k
k=0
and so
t
ta (t) − 4ta(t) + 2 a(s)ds = 0,
0
1
L[a(t)]( p) = . (2.9)
p2 − 4p
Proof One possibility to compute the Laplace transform of a(t) is to start from (2.8):
∞ ∞
1 2k 1 2k k!
L[a(t)]( p) = L[t k ]( p) =
k! k k! k p k+1
k=0 k=0
∞
1 2k 1 1 1
= k
= ,
p k p p 1− 4
k=0 p
so that
1
L[a(t)]( p) = .
p2 − 4p
The same result can be obtained if we use (2.7). Indeed, let A( p) = L[a(t)]( p).
Then
Applying L to (2.7) and using the above equations we get the following differential
equation for A( p):
(4 p − p 2 )A ( p) + (2 − p)A( p) = 0,
C
with the general solution A( p) = .
− 4p p2
From the theory of Laplace transform we know that lim p A( p) = a(0) = 1, so
p→∞
C = 1.
2.3 The generating function a(t) and the Bessel function I0 (t)
1 1
L[a(t)]( p) = √ √
p p−4
we see that
4t
1 e
L[a(t)]( p) = L √ ( p) · L √ ( p).
πt πt
Therefore,
t e4s 1
a(t) = √ √ ds. (2.10)
0 π s π(t − s)
On the other hand, it is known (see, e.g., [2, (9.6.16)]) that the modified Bessel
function of first kind of order zero can be represented as
π
1
I0 (t) = e−t cos s ds, t ∈ R.
π 0
A second proof of this theorem runs as follows. It is known (see, e.g., [2, (9.6.10)]))
that
∞
2k
1 t
I0 (t) = , t ∈ R. (2.14)
(k!)2 2
k=0
and w(0) = 1, w (0) = 2. This fact in conjunction with Theorem 2.1 shows that
w(t) = a(t), i.e., (2.13) holds.
which is the expression of a(t) furnished by the integral (4) in [21]. Theorem 2.4 is
obtained using either (2.10) (as above) or the integral (4) (as in [21]), combined with
suitable changes of variables.
Let us remark that (2.10) is obtained using the Laplace transform while [21, (4)] is
a consequence of other classical formulas. On the other hand, Theorem 2.4 is given
by Michael Somos without proof in [22, A000984].
We are indebted to a reviewer for suggesting these connections. Another reviewer
remarked that one can give a direct series proof of (2.13) as follows
∞
∞ ∞ [n/2]
2l 2k+l 1 n n 2k n−2k
e I0 (2t) =
2t
t = t 2
(k!) l!
2 n! 2k k
k=0 l=0 n=0 k=0
∞
1 n 2n
= t = a(t),
n! n
n=0
[n/2]
n
2k n−2k
2n
2 =
2k k n
k=0
2.4 The generating function and the index of coincidence for Poisson distribution
tk
pk (t) := e−t , k = 0, 1, 2, . . . .
k!
The index of coincidence associated with this distribution is (see, e.g., [16] and the
references therein)
∞
∞
t 2k
K 1 (t) := pk2 (t) = e−2t . (2.16)
(k!)2
k=0 k=0
It is known (see [17] and the references therein) that K 1 is a logarithmically convex
function on [0, ∞). Since log a(t) = 4t + log K 1 (t), we have
Theorem 2.5 The generating function a(t) is log-convex on [0, ∞). Moreover, using
the bounds for K 1 established in [4], it follows that for t ≥ 0,
√
√ 1/2
2e 1+4t −1−2t 2e 1+16t −1−4t
2 2
a(t) 1
√ ≤ 4t ≤ √ ≤√ .
1 + 4t + 1
2 e 1 + 16t 2 + 1 4t + 1
Using (2.3) for the sequence of Appell polynomials (Un (x))n≥0 , we can write
∞
1
a(t)etnx = Uk (nx)t k , n = 0, 1, 2, . . . . (3.1)
k!
k=0
The Jakimovski–Leviatan type operators associated with (Un (x))n≥0 are defined by
∞
e−nx 1 k
n f (x) := Uk (nx) f , n ≥ 1, (3.2)
a(1) k! n
k=0
where f ∈ C[0, ∞) is a function for which the series is convergent for all x ∈ [0, ∞).
Let e j (x) = x j , x ∈ [0, ∞), j = 0, 1, 2, . . . .
∞
1
From (3.1) we get a(1)enx = Uk (nx); combined with (3.2) for f = e0 , this
k!
k=0
yields
n e0 = e0 , n ≥ 1. (3.3)
In particular, for t = 1,
∞
1
a (1)enx + a(1)nxenx = Uk (nx)k,
k!
k=0
which leads to
∞
e−nx 1 k 1 a (1)
Uk (nx) = x + .
a(1) k! n n a(1)
k=0
Therefore,
1 a (1)
n (e1 ) = e1 + e0 , n ≥ 1. (3.4)
n a(1)
and so on.
a (1) a (1) a ( j) (1)
Apparently, the explicit expression of n e j depends on , ,..., .
a(1) a(1) a(1)
In fact, we have
a (1)
Theorem 3.1 The explicit expression of n e j depends only on .
a(1)
a ( j) (1)
Set r j := , j ≥ 0. From (3.7) with t = 1 we get
a(1)
a (1)
It follows that r2 , r3 , . . . can be expressed in terms of r0 = 1 and r1 = , and this
a(1)
concludes the proof.
a (1) a (1)
=3 + 2.
a(1) a(1)
j j
j
s( j, i)n i n ei = r j−i n i ei , (3.9)
i
i=0 i=0
j
∞
j (i) 1
a (t)(nx) j−i etnx = Uk (nx)k(k − 1) . . . (k − j + 1)t k− j .
i k!
i=0 k=0
j (i)
∞
e−nx 1
j
j a (1)
(nx) j−1
= Uk (nx) s( j, i)k i ,
i a(1) a(1) k!
i=0 k=0 i=0
and therefore
j
∞ i
e−nx 1
j
j k
ri (nx) j−i = s( j, i)n i Uk (nx) .
i a(1) k! n
i=0 i=0 k=0
This implies
j j
j
s( j, i)n n ei (x) =
i
ri (nx) j−i , x ≥ 0,
i
i=0 i=0
which is (3.9).
j
j−1
j
n e j = r j−i n ei −
i− j
s( j, i)n i− j n ei , (3.10)
i
i=0 i=0
Proof Let λ ∈ R, λ < 0, and ϕ(x) := eλx , x ∈ [0, ∞). Then, according to (3.2),
∞
e−nx 1 λ
k
n ϕ(x) = Uk (nx) e n .
a(1) k!
k=0
λ
Using (3.1) with t := e n we get
e−nx λ λ
n ϕ(x) = a(e n )enxe , i.e.,
n
a(1)
λ λ
a(e n ) nx e n −1
n ϕ(x) = e .
a(1)
This implies
lim n ϕ = ϕ. (3.12)
n→∞
Now (3.11) is a consequence of (3.12) and (3.3), combined with [3, Proposition 4.2.5
(7)].
∞ j+1
ne−nx 1 n
K n f (x) := U j (nx) f (t)dt, f ∈ C ∗ [0, ∞), x ∈ [0, ∞).
a(1) j! j
j=0 n
lim K n f = f .
n→∞
Proof As in the proof of Theorem 3.3, it is easy to verify that for λ < 0 and ϕ(x) = eλx ,
n λ
K n ϕ(x) = e n − 1 n ϕ(x).
λ
lim K n ϕ = ϕ. (3.13)
n→∞
An application of (3.3), (3.13) and [3, Proposition 4.2.5 (7)] concludes the proof.
Acknowledgements We are very grateful to the reviewers for their highly valuable comments and sugges-
tions which improved considerably the first version of the paper.
This work was supported by a Hasso Plattner Excellence Research Grant (LBUS-HPI-ERG-2020-04),
financed by the Knowledge Transfer Center of the Lucian Blaga University of Sibiu.
Data Availability Data sharing not applicable to this article as no datasets were generated or analysed during
the current study.
References
1. Abel, U., Ivan, M.: Asymptotic expansion of the Jakimovski–Leviatan operators and their derivatives.
In: Functions, Series, Operators(Budapest, 1999), János Bolyai Math. Soc., Budapest, pp. 103–119
(2002)
2. Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions with Formulas, Graphs, and
Mathematical Tables. National Bureau of Standards, New York (1972)
3. Altomare, F., Campiti, M.: Korovkin-type Approximation Theory and Its Applications, de Gruyter
Studies in Mathematics 17. Walter de Gruyter, Berlin (1994)
4. Acu, A.M., Bascanbaz-Tunca, G., Raşa, I.: Bounds for indices of coincidence and entropies. Math.
Inequl. Appl. (in press)
5. Bretti, G., Cesarano, C., Ricci, P.E.: Laguerre-type exponentials and generalized Appell polynomials.
Comput. Math. Appl. 48, 833–839 (2004)
6. Bretti, G., Ricci, P.E.: Multidimensional extensions of the Bernoulli and Appell polynomials. Taiwan
J. Math. 8(3), 415–428 (2004)
7. Costabile, F.A., Longo, E.: A determinantal approach to Appell polynomials. J. Comput. Appl. Math.
234(5), 1528–1542 (2010)
8. Douak, K.: The relation of the d-orthogonal polynomials to the Appell polynomials. J. Comput. Appl.
Math. 70, 279–295 (1996)
9. Graham, R., Knuth, D., Patashnik, O.: Concrete Mathematics: A Foundation for Computer Science.
Addison-Wesley, Boston (1994)
10. Gupta, V., Rassias, M.T.: Moments of Linear Positive Operators and Approximation. Springer, New
York (2019)
11. Gupta, P., Acu, A.M., Agrawal, P.N.: Jakimovski-Leviatan operators of Kantorovich type involving
multiple Appell polynomials. Georgian Math. J. 28(1), 73–82 (2021)
12. He, M.X., Ricci, P.E.: Differential equation of Appell polynomials via the factorization method. J.
Comput. Appl. Math. 139, 231–237 (2002)
13. Ismail, M.E.H.: Remarks on differential equation of Appell polynomials via the factorization method.
J. Comput. Appl. Math. 154, 243–245 (2003)
14. Jakimovski, A., Leviatan, D.: Generalized Szász operators for the approximation in the infinite interval.
Mathematica (Cluj) 34, 97–103 (1969)
15. Mursaleen, M., Al-Abied, A., Acu, A.M.: Approximation by Chlodowsky type of Szasz operators
based on Boas-Buck-type polynomials. Turk. J. Math. 42, 2243–2259 (2018)
16. Raşa, I.: Entropies and Heun functions associated with positive linear operators. Appl. Math. Comput.
268, 422–431 (2015)
17. Raşa, I.: Convexity properties of some entropies. Results Math. 73, 105 (2018)
18. Raşa, I.: Convexity properties of some entropies (II). Results Math. 74, 154 (2019)
19. Sucu, S., Içöz, G., Varma, S.: On some extensions of Szász operators including Boas-Buck-type
polynomials. Abstract Appl Anal; Article ID 680340, p. 15 (2012)
20. Varma, S., Sucu, S., Içöz, G.: Generalization of Szász operators involving Brenke type polynomials.
Comput. Math. Appl. 64, 121–127 (2012)
21. https://math.stackexchange.com/questions/3210839/the-exponential-generating-function-for-the-cen
tral-binomial-coefficients
22. https://oeis.org/
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
1. use such content for the purpose of providing other users with access on a regular or large scale
basis or as a means to circumvent access control;
2. use such content where to do so would be considered a criminal or statutory offence in any
jurisdiction, or gives rise to civil liability, or is otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association
unless explicitly agreed to by Springer Nature in writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a
systematic database of Springer Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a
product or service that creates revenue, royalties, rent or income from our content or its inclusion as
part of a paid for service or for other commercial gain. Springer Nature journal content cannot be
used for inter-library loans and librarians may not upload Springer Nature journal content on a large
scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not
obligated to publish any information or content on this website and may remove it or features or
functionality at our sole discretion, at any time with or without notice. Springer Nature may revoke
this licence to you at any time and remove access to any copies of the Springer Nature journal content
which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or
guarantees to Users, either express or implied with respect to the Springer nature journal content and
all parties disclaim and waive any implied warranties or warranties imposed by law, including
merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published
by Springer Nature that may be licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a
regular basis or in any other manner not expressly permitted by these Terms, please contact Springer
Nature at
onlineservice@springernature.com