mathematics

Article
Extending the Applicability of Stirling’s Method
Cristina Amorós 1 , Ioannis K. Argyros 2 , Á. Alberto Magreñán 3 , Samundra Regmi 2 ,
Rubén González 1, * and Juan Antonio Sicilia 1
1 Escuela Superior de Ingeniería y Tecnología, Universidad Internacional de La Rioja, 26006 Logroño, Spain;
[email protected] (C.A.); [email protected] (J.A.S.)
2 Department of Mathematics Sciences, Cameron University, Lawton, OK 73505, USA;
[email protected] (I.K.A.); [email protected] (S.R.)
3 Departamento de Matemáticas y Computación, Universidad de La Rioja, 26004 Logroño, Spain;
[email protected]
* Correspondence: [email protected]




Received: 30 November 2019; Accepted: 28 December 2019; Published: 31 December 2019


Abstract: Stirling’s method is considered as an alternative to Newton’s method when the latter fails
to converge to a solution of a nonlinear equation. Both methods converge quadratically under similar
convergence criteria and require the same computational effort. However, Stirling’s method has
shortcomings too. In particular, contractive conditions are assumed to show convergence. However,
these conditions limit its applicability. The novelty of our paper lies in the fact that our convergence
criteria do not require contractive conditions. Hence, we extend its applicability of Stirling’s method.
Numerical examples illustrate our new findings.

Keywords: Stirling’s method; Newton’s method; convergence; Fréchet derivative; banach space

1. Introduction
In this work we deal with finding a fixed point x ∗ of the equation

x = F ( x ), (1)

where F is a Fréchet-differentiable operator defined on a convex subset D of a Banach space X with

values into itself. By I we denote the identity linear operator in L( X, X ). The symbol L( X, X ) stands
for the space of bounded linear operators from X into X.
Many applications from different areas, including education, reduce to dealing with Equation (1)
utilizing mathematical modelling [1–24]. However, the solution x ∗ is found in closed form only in rare
cases. This problem leads to the usage of methods that are iterative in nature.
We study Stirling’s method given for all n = 0, 1, 2, . . . by

xn+1 = xn − ( I − F 0 ( F ( xn )))−1 ( xn − F ( xn )), (2)

where x0 ∈ D. Further we will introduce an operator Γ( x ) ∈ L( X, X ) such that Γ( x ) = ( I − F 0 ( F ( x )))−1

with x ∈ D, and denote Γ0 = Γ( x0 ) for use in later Sections.
This method converges quadratically as Newton’s method does, and also requires the same
computational effort (see details in [1,22]). It is considered to be a useful alternative in cases where
Newton’s method fails to converge (see such examples in [22]). However, the usage of Stirling’s
method has a drawback, since the convergence criteria require contractions. We have detected some
other problems listed in Remarks 3 and 4. These drawbacks limit the applicability of Stirling’s method.

Mathematics 2020, 8, 35; doi:10.3390/math8010035 www.mdpi.com/journal/mathematics

Mathematics 2020, 8, 35 2 of 10

In order to extend its applicability, we do not use contractive conditions in our semi-local as well as
the local convergence results.
The rest of the work is structured as follows. Section 2 includes the semi-local convergence
analysis. Section 3 contains the local analysis. The numerical results are given in Section 4.

2. Semi-Local Convergence Analysis

Let L0 > 0, L > 0 and γ ≥ 0. Consider a real sequence {tn } as

L ( t n +1 − t n )2
t0 = 0, t1 = γ, tn+2 = tn+1 + . (3)
2 (1 − L 0 t n +1 )

Next, we study the convergence of sequence {tn } by developing relevant lemmas and theorems.

Lemma 1. Suppose that

1
h = L1 γ < , (4)
2
where
1 p
L1 = ( L + 4L0 + L2 + 8L0 L).
8
Then, sequence {tn } generated for t0 = 0 by (4) is increasing, converges to its unique least upper bound t∗ ,
so that

d1 ≤ t ∗ ≤ d2 , (5)

where
L0 γ
1 − exp[− (1− L ]
0 γ )(1− δL0 γ )
d1 =
L0
1 − exp[ 22L 0γ
− L0 γ +
2δ1
2−δ1 ]
d2 = ,
L0
L
δ=
2L0 (1 − L0 γ)2
and
L δ 2
δ1 = ( ) .
2L0 1 − δ

Proof. It is convenient to first simplify sequence {tn }. Define sequence {αn } by αn = 1 − L0 tn .

L ( α n − α n −1 )2
Then, by (4) we can write α0 = 1, α1 = 1 − L0 γ, αn+1 = αn − 2L0 αn . Moreover, define sequence
L θn 2
{θn } by θn = 1 − ααn−n 1 . Then, we can write θ1 = L0 γ, θn+1 = 2L0 ( 1−θn ) . We have by (4) that δθ1 < 1
and 0 < θ2 < θ1 . Suppose that 0 < θk < θk−1 and δθk < 1. Then, we get in turn that

L θ
θ k +1 = ( k )2 < δθk2 < θk (6)
2L0 1 − θk
and
δθk+1 < δθk < 1. (7)
1− α n
Hence, { β n } is a decreasing sequence, so αn = (1 − β n )αn−1 and tn = L0 are also decreasing
sequences. In particular,

α n = (1 − β n ) α n −1 = . . . = (1 − β n ) . . . (1 − β 1 ) α 0 = (1 − β n ) . . . (1 − β 0 ).
Mathematics 2020, 8, 35 3 of 10

1− α n
From 0 < β 1 = L0 γ < 1, we get 0 < αn < 1, so tn = γ < γ1 . That is sequence {tn } is increasing,
bounded from above by L10 , so it converges to t∗ .
Next, we show (4). We can write
∞
α∗ = lim αn =
n→∞
∏ (1 − β n ),
n =1

or
∞
1 1
log
α ∗
= ∑ log
1 − βn
.
n =1

Using the estimate

t−1 t2 − 1
2 ≤ log t ≤ for t > 1,
t+1 2t
we get first an upper bound for log α1∗ by (5) and (6) and the inequality 2n ≥ n + 1 for n = 0, 1, 2 . . .:

∞ ∞
1 β n (2 − β n ) 1
log
α∗
≤ ∑ 2(1 − β n )
≤
1 − β1 ∑ β n +1
n =1 n =0
∞ ∞
1 1
≤
δ (1 − β 1 ) ∑ (δθ1 ) 2n
≤
δ (1 − β 1 ) ∑ (δβ1 )n
n =1 n =1
β1
= ,
(1 − β 1 )(1 − δβ 1 )

1− α ∗
which together with t∗ = L0 imply t∗ ≤ d2 . The lower bound in (4) is obtained similarly from
the estimate:
∞
1 αn 2α1 2α2
log
α ∗
≥ 2 ∑ 2 − αn
>
2 − α1
+
2 − α2
.
n =1

Lemma 2. Suppose that

1
h= . (8)
2
1
Then, sequence {tn } is increasingly converging to L0 .

1−(1− L0 γ)n
Proof. We have αn = (1 − L0 R)n , β n = L0 γ and tn = L0 . Then, by (8), we get 0 ≤ L0 γ < 1.

In what follows the set denoted by U ( x, a) is a ball with center x ∈ X and of radius a > 0.
To simplify, the notation, by || || in this work, we denote the operator norm or the norm on the
Banach space. The semi-local convergence analysis is based on the conditions (C ):
(C1 ) F : D ⊂ X → X is a Fréchet differentiable operator and there exist x0 ∈ D, c > 0, γ ≥ 0 such
that Γ0 = ( I − F 0 ( F ( x0 )))−1 ∈ L( X, X ) with

|| I − F 0 ( F ( x0 ))|| ≤ c

and
||Γ0 ( x0 − F ( x0 ))|| ≤ γ.

(C2 ) There exist a0 ∈ [0, 1), b0 > 0 such that for each x ∈ D

|| F ( x ) − F ( x0 )|| ≤ a0 || x − x0 ||
Mathematics 2020, 8, 35 4 of 10

and
||Γ0 ( F 0 ( F ( x )) − F 0 ( F ( x0 ))|| ≤ b0 || F ( x ) − F ( x0 )||.
1
(C3 ) Let r0 = a0 b0 and D0 = D ∩ U ( x0 , r0 ). There exist b > 0, b1 > 0 such that for each x, y ∈ D0

||Γ0 ( F 0 ( x ) − F (y))|| ≤ b|| x − y||

and
|| F 0 ( F ( x )) − F 0 ( F ( x0 ))|| ≤ b1 || F ( x ) − F ( x0 )||.

(C4 ) Hypotheses of Lemmas 1 and 2 hold with

b1 1
L = 2b(c + + )
b0 2

and
L0 = a0 b0 .

(C5 )
|| F ( x0 ) − x0 ||
≤ t∗ .
1 − a0
(C6 ) The ball Ū ( x0 , t∗ ) is constructed such that

Ū ( x0 , t∗ ) ⊆ D.

We suppose from now on that the conditions (C ) hold.

Next, the semi-local convergence result is given for Stirling’s method (2).

Theorem 1. Under conditions (C ), sequence { xn } generated by Stirling’s method (2) is well defined, remains
in U ( x0 , t∗ ) for each n = 0, 1, 2, . . . and converges to x ∗ ∈ Ū ( x0 , t∗ ) which satisfies x ∗ = F ( x ∗ ) with Q-order
of convergence 2. Moreover, the following estimates hold

|| xn − x ∗ || ≤ t∗ − tn ,

and x ∗ is the only fixed point of F in U ( x0 , t∗∗ ), with

2
t∗∗ = − (2a0 + 1)t∗ .
b

Proof. Let x ∈ Ū ( x0 , t∗ ). We get by (C2 ) and (C5 ) that

|| F ( x ) − x0 || ≤ || F ( x ) − F ( x0 )|| + || F ( x0 ) − x0 || ≤ a0 || x − x0 || + || F ( x0 ) − x0 || ≤ a0 t∗ + || F ( x0 ) − x0 || ≤ t∗ ,

so F ( x ) ∈ Ū ( x0 , t∗ ). Using (C2 ) and the Lemmas 1 and 2, we have in turn that

||Γ0 (Γ( x ) − Γ0 )|| = ||Γ0 (( F 0 ( F ( x )) − F 0 ( F ( x0 )))|| (9)

∗
≤ b0 a0 || x − x0 || = L0 || x − x0 || ≤ L0 t < 1.

By the Lemma of Banach on invertible operators [21] (Perturbation Lemma 2.3.2, p. 45) Γ( x )−1 ∈
L(X , X ), and
1
||Γ( x )( I − F 0 ( F ( x0 )))|| ≤ . (10)
1 − L0 || x − x0 ||
Using Stirling’s method (2):
Mathematics 2020, 8, 35 5 of 10

x k +1 − F ( x k +1 ) = F ( x k +1 ) − F ( x k ) − F ( x k +1 ) + F ( x k ) (11)
0
= F (yk )( xk+1 − xk ) − ( F ( xk+1 − F ( xk ))
Z 1
= [ F 0 (yk ) − F 0 ( xk + θ ( xk+1 − xk ))]( xk+1 − xk )dθ.
0

Then, in view of (C2 ), (C3 ) and Equation (11), we obtain in turn that

||Γ0 ( xk+1 − F ( xk+1 ))|| (12)

Z 1
≤ b ||yk − xk − θ ( xk+1 − xk )|||| xk+1 − xk ||dθ
0
1
≤ b[||yk − xk || + || xk+1 − xk ||]|| xk+1 − xk ||
2
1
≤ b[(|| I − F 0 ( F ( x0 ))|| + || F 0 (yk ) − F 0 ( F ( x0 ))||) + || xk+1 − xk || + || xk+1 − xk ||]|| xk+1 − xk ||
2
b a 1
≤ b(c + 1 0 + )|| xk+1 − xk ||2
a0 b0 2
L
= || x − xk ||2 .
2 k +1

Next, we can connect the preceding estimates on sequence { xk } with {tk }. Indeed, we get by (C1 )
and Equation (3) that

|| x1 − x0 || = ||Γ0 ( x0 − F ( x0 ))|| ≤ γ = t1 = t1 − t0 .

By induction, Equations (3), (4), (10) and (12), we have in turn that

|| xk+1 − xk || = ||Γk ( xk − F ( xk ))|| ≤ ||Γk ( I − F 0 ( F ( x0 )))||||Γ0 ( xk − F ( xk ))|| (13)

L ( t k − t k −1 )2
≤ = t k +1 − t k .
2(1 − L0 t k )

Hence, {tk } defined by Equation (3) is a majorizing sequence for { xk }. By Lemmas 1 and 2,
sequence {tk } is complete as convergent to t∗ . It then follows by Equation (13) that sequence { xk } is
also complete so it converges to some x ∗ ∈ Ū ( x0 , t∗ ). By the estimate (see (12))

L L
||Γ0 ( xk+1 − F ( xk+1 ))|| ≤ || xk+1 − xk ||2 ≤ (tk+1 − tk )2 , (14)
2 2

we deduce that x ∗ = F ( x ∗ ) by letting k → ∞. Estimate || xn − x ∗ || ≤ t∗ − tn follows from Equation (13)

and for λ = 2(1−LL t∗ ) , we get that
0

L
|| xk+1 − xk || ≤ || x − xk−1 ||2
2(1 − L0 t k ) k
≤ λ|| xk − xk−1 ||2 ,
Mathematics 2020, 8, 35 6 of 10

which implies that the Q-order convergence of Stirling’s method (2) is two. Furthermore, to show the
R1
uniqueness part, let y∗ ∈ U ( x0 , t∗∗ ) with F (y∗ ) = y∗ . Define the operator Q by Q = − 0 Γ0 F 0 ( x ∗ +
θ (y∗ − x ∗ ))dθ. In view of (C2 ) and (C3 ), we obtain in turn that
Z 1
|| I − (Γ0 − Q)|| = || Γ0 [ F 0 ( x ∗ + θ (y∗ − x ∗ )) − F 0 ( F ( x0 ))]dθ || (15)
0
Z 1
≤ b || x ∗ + θ (y∗ − x ∗ ) − F ( x0 )||dθ
0
1 1
≤ b[|| F ( x ∗ ) − F ( x0 )|| + || x ∗ − x0 || + ||y∗ − x0 ||]
2 2
1 1
≤ b[( a0 + )t∗ + t∗∗ ] < 1.
2 2

Then, by (15) (Γ0 − Q)−1 ∈ L( X, X ). Finally, we obtain y∗ = x ∗ using the identity

0 = Γ0 (y∗ − F (y∗ ) − x ∗ + F ( x ∗ )) = (Γ0 − Q)(y∗ − x ∗ ).

Remark 1.

(a) The Stirling’s method usual conditions corresponding to (C2 ) (first condition) are given by [22]:

(C2 )’ || F 0 ( x )|| ≤ a for each x ∈ D and a ∈ [0, 1).

That is, operator F must be a contraction on D. Moreover, the convergence of Stirling’s method was shown
in [22] under (C2 ), D0 = D and a ∈ (0, 13 ]. However, in the present study no such assumption is made.
Hence, the applicability of Stirling’s method (2) is extended. Notice also that we can have a0 ≤ a, b0 ≤ b
and c can be chosen as b = cb1 .
(b) Estimate (4) is similar to the sufficient convergence Kantorovich-type criteria for the semi-local convergence
of Newton’s method given by us in [4]. However, the constants b¯0 and b̄ are the center-Lipschitz and
Lipschitz constants for operator F (see also part (e)).
(c) If set D0 is switched by D1 = D ∩ U ( x1 , r0 − || x0 − F ( x0 )||), since D1 ⊆ D and the iterates remain
in D1 the results can be improved even further. The corresponding constants to b and b1 will be at least
as small.
(d) In view of the proof of Theorem 1, scalar sequence {sn } defined by

k n ( s n − s n −1 )2
s0 = 0, s1 = R, sn+1 = sn + ,
1 − L0 s n

is also a majorizing sequence for Stirling’s method (2), where

1
k n = ab(c + b1 a0 sn + ) < L
2
sn ≤ tn ,

s n +1 − s n ≤ t n +1 − t n

and
s∗ = lim sn ≤ t∗ .
n→∞

(e) Newton’s method for Equation (1) is given for all n = 0, 1, 2, . . . by

yn+1 = yn − ( I − F 0 (yn ))−1 (yn − F (yn )).

Mathematics 2020, 8, 35 7 of 10

Consider, items c̄, γ̄, L¯0 , L̄, L¯1 , Γ¯0 , b¯0 , b̄, b¯1 , r¯0 , D̄0 and h̄, corresponding to c, γ, L0 , L, L1 , Γ0 , b0 , b, b1 ,
r0 , D0 and h respectively as
|| I − F 0 ( x0 )|| ≤ c̄,
||Γ¯0 ( x0 − F ( x0 ))|| ≤ γ̄,
||Γ¯0 ( F 0 ( x ) − F 0 ( x0 ))|| ≤ b¯0 || x − x0 ||,
||Γ¯0 ( F 0 ( x ) − F 0 (y))|| ≤ b̄|| x − y||,
b¯ 1
L̄ = 2b̄(c̄ + ¯1 + ), b¯1 = b¯0 ,
b0 2
L¯0 = a¯0 b¯0 ,
1
r¯0 = ,
a¯0 b¯0
D̄0 = D ∩ U ( x0 , r¯0 ),

and
1
h̄ = L¯1 , γ̄ ≤ ,
2
where
1 p
L¯1 = ( L̄ + 4 L¯0 + L̄2 + 8 L¯0 L̄).
8

The scalar sequence t̄n is defined as

L̄(t̄n+1 − t̄n )2
t¯0 = 0, t¯1 = γ̄, t̄n+2 = t̄n+1 +
2(1 − L¯0 t̄n+1 )

Then, Stirling’s method sufficient convergence criteria, error bounds and information on the uniqueness of
the solution are better than Newton’s method when the "bar" constants and sets are smaller than the non
bar constants. Similar favorable comparison can be made in the local convergence case that follows.

3. Local Convergence
The conditions (H) are used in the local convergence analysis of Stirling’s method (2):

(H1 ) F : D ⊂ X → X is a Fréchet differentiable operator, and there exists x ∗ ∈ D such that

Γ∗ = ( I − F 0 ( x ∗ ))−1 ∈ L( X, X ) and F ( x ∗ ) = x ∗ .
(H2 ) There exist µ ∈ (0, 1), ξ 0 > 0 such that for each x ∈ D

|| F ( x ) − F ( x ∗ )|| ≤ µ|| x − x ∗ ||

and
||Γ∗ ( F 0 ( F ( x )) − F 0 ( F ( x ∗ )))|| ≤ ξ 0 || F ( x ) − F ( x ∗ )||.

(H3 ) Let D0∗ = D ∩ U ( x ∗ , R0 ), R0 = 1

ξ0 µ . There exists ξ > 0 such that for each x, y ∈ D0∗

||Γ∗ ( F 0 ( x ) − F 0 (y))|| ≤ ξ || x − y||.

(H4 ) The ball Ū ( x ∗ , R) is constructed such that Ū ( x ∗ , R) ⊆ D, where

1
R= .
(µ + 12 )ξ + µξ 0
Mathematics 2020, 8, 35 8 of 10

Theorem 2. Suppose that conditions (H) hold. Then, sequence { xn } generated for x0 ∈ U ( x ∗ , R) − { x ∗ } by
Stirling’s method (2) is well defined in U ( x ∗ , R), remains in U ( x ∗ , R) for each n = 0, 1, 2, . . . and converges to
x ∗ ∈ Ū ( x ∗ , R). Moreover, the following inequality holds

ξ (µ + 12 )|| xn − x ∗ ||2
|| xn+1 − x ∗ || ≤ . (16)
1 − µξ 0 || xn − x ∗ ||

Furthermore, if R1 = 2ξ , x ∗ is the only fixed point of F on U ( x ∗ , R1 ).

Proof. We shall show using mathematical induction that sequence { xn } is well defined, remains in
U ( x ∗ , R) and converges to x ∗ so that (16) is satisfied. We have by (H1 ) and (H2 ) for x ∈ U ( x ∗ , R) that

|| F ( x ) − x ∗ || = || F ( x ) − F ( x ∗ )|| ≤ µ|| x − x ∗ || ≤ R,

so F ( x ) ∈ U ( x ∗ , R). Then by (H2 )

||Γ∗ ( I − F 0 ( F ( x ))) − Γ∗ || = ||Γ∗ ( F 0 ( F ( x )) − F 0 ( F ( x ∗ )))|| (17)

∗ ∗
≤ ξ 0 || F ( x ) − F ( x )|| ≤ ξ 0 µ|| x − x || ≤ ξ 0 µR < 1.

Hence, Γ( x ) ∈ L( X, X ) and

1
||Γ( x )( I − F ( F ( x ∗ )))|| ≤ . (18)
1 − µ0 ξ 0 || x − x ∗ ||

In particular, (18) holds for x = x0 , which shows that x1 is well defined by Stirling’s method for
n = 0. We can write by (H1 ) that

x1 − x ∗ = x0 − x ∗ − ( I − F 0 ( F ( x0 )))−1 ( x0 − F ( x0 )) (19)
0 −1 ∗ 0 ∗
= ( I − F ( F ( x0 ))) [ F ( x0 ) − F ( x ) − F ( F ( x0 ))( x0 − x )]
Z 1
0 −1
= ( I − F ( F ( x0 ))) [ ( F 0 ( x ∗ + θ ( x0 − x ∗ )) − F 0 ( F ( x0 ))( x0 − x ∗ )dθ ].
0

We get in turn by (H2 ) and (H3 )

Z 1
||Γ∗ ( F 0 ( x ∗ + θ ( x0 − x ∗ )) − F 0 ( F ( x0 )))( x0 − x ∗ )dθ || (20)
0
Z 1
≤ ξ || x ∗ + θ ( x0 − x ∗ ) − F ( x0 )|||| x0 − x ∗ ||dθ
0
≤ ξ [|| F ( x ∗ ) − F ( x0 )|| + θ || x0 − x ∗ ||]dθ
1
≤ ξ (µ + )|| x0 − x ∗ ||
2

Then, by (18)–(20), we get that also

|| x1 − x ∗ || ≤ ||( I − F 0 ( F ( x0 )))−1 Γ∗ ||

ξ (µ + 21 )|| x0 − x ∗ ||2
≤ ≤ || x0 − x ∗ || < R,
1 − µξ 0 || x0 − x ∗ ||
Mathematics 2020, 8, 35 9 of 10

so (16) holds for n = 0 and x1 ∈ U ( x ∗ , R). Switch x0 by xk in the preceding estimates, we arrive at
(16). In view of the estimate || xk+1 − x ∗ || < || xk − x ∗ || < R, we conclude that limk→∞ xk = x ∗ and
xk+1 ∈ U ( x ∗ , R). Let x0 = x ∗ in (15) to show the uniqueness part.

Remark 2. The local results in the literature use (C2 )’ and D0∗ = D. But (H2 ) is weaker than (C2 )’.
Hence, we extend the applicability of Stirling’s method (2) in the local case too.

4. Numerical Example with Concluding Remarks

In the next example, we compare Stirling’s method with Newton’s method.

Example 1. Let D = X = R. Consider function F on D as


1
 − 3 x,
 x≤3
1 2 7
F(x) = 3 x − 3 x + 3, 3≤x≤4
 1 ( x − 7),

x ≥ 4.
3

Clearly, the quadratic polynomial joins smoothly with the linear parts.

(I) Semilocal case (i). If we choose x0 = 3, we see that x1 = y1 = x ∗ = 0. Moreover, the semi-local
convergence criteria of Theorem 1 are satisfied (with γ = 0, a0 = 31 and c = 43 ).
(II) Local convergence criteria of Theorem 2 (with µ = 13 , since the derivative of the quadratic
polynomial satisfies 31 |2x − 7| ≤ 13 for all x ∈ [3, 4]).
(III) In Tables 1 and 2 we present some cases in which Stirling’s method stands better than
Newton’s one.

Table 1. Iteration of Newton’s and Stirling’s method with different starting points.

Iteration Newton’s Method Stirling’s Method

0 3.4975 3.4975
1 −646.501 0.0618766
2 −1.13687 × 10−13 0

Table 2. Iteration of Newton’s and Stirling’s method with different starting points.

Iteration Newton’s Method Stirling’s Method

0 3.5 3.5
1 ∞ 0.0625
2 − 0

In the current study, we have successfully demonstrated our claims on Stirling’s method by
focusing on very classic problems, but in the future we will consider studying other complex problems
such us solving symmetric ordinary differential equations with a more favorable theory.

Author Contributions: All authors have equally contributed to this work. All authors have read and agreed to
the published version of the manuscript.
Funding: Research supported in part by Séneca 20928/PI/18 and by MINECO PGC2018-095896-B-C21.
Conflicts of Interest: The authors have no conflict of interest. The funders had no role in the design of the study;
in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish
the results.
Mathematics 2020, 8, 35 10 of 10

References
1. Argyros, I.K. Stirling’s method and fixed points of nonlinear operator equations in Banach spaces. Bull. Inst.
Math. Acad. Sin. 1995, 23, 13–20.
√
2. IArgyros, I.K. A new iterative method of asymptotic order 1 + 2 for the computation of fixed points. Int. J.
Comput. Math. 2005, 82, 1413–1428. [CrossRef]
3. Argyros, I.K.; Chen, D. Results on the Chebyshev method in Banach spaces. Proyecciones 1993, 12, 119–128.
[CrossRef]
4. Argyros, I.K.; Hilout, S. Weaker conditions for the convergence of Newton’s method. J. Complex. 2012,
28, 364–387. [CrossRef]
5. Argyros, I.K. A unifying local–semilocal convergence analysis and applications for two–point Newton–like
methods in Banach space. J. Math. Anal. Appl. 2004, 298, 374–397. [CrossRef]
6. Argyros, I.K.; Magreñán, Á.A. Iterative Methods and Their Dynamics with Applications: A Contemporary Study;
CRC Press: Boca Ratón, FL, USA, 2017.
7. Burrows, A.; Lockwood, M.; Borowczak, M.; Janak, E.; Barber, B. Integrated STEM: Focus on Informal
Education and Community Collaboration through Engineering. Educ. Sci. 2018, 8, 4. [CrossRef]
8. Cianciaruso, F.; De Pascale, E. Estimates of majorizing sequences in the Newton-Kantorovich method:
A further improvement. J. Math. Anal. Appl. 2006, 332, 329–335. [CrossRef]
9. Cordero, A.; Gutiérrez, J.M.; Magreñán, Á.A.; Torregrosa, J.R. Stability analysis of a parametric family of
iterative methods for solving nonlinear models. Appl. Math. Comput. 2016, 285, 26–40. [CrossRef]
10. Chun, C.; Stanica, P.; Neta, B. Third order family of methods in Banach spaces. Comput. Math. Appl. 2011,
61, 1665–1675. [CrossRef]
11. Grout, I. Remote Laboratories as a Means to Widen Participation in STEM Education. Educ. Sci. 2017, 7, 85.
[CrossRef]
12. Hernández, M.Á. The Newton Method for Operators with H’older Continuous First Derivate. J. Optim.
Theory Appl. 2001, 109, 631–648. [CrossRef]
13. Kantorovich, L.V.; Akilov, G.P. Functional Analysis; Pergamon Press: Oxford, UK, 1982.
14. Laforce, M.; Noble, E.; Blackwell, C. Problem-Based Learning (PBL) and Student Interest in STEM Careers:
The Roles of Motivation and Ability Beliefs. Educ. Sci. 2017, 7, 92. [CrossRef]
15. LeTendre, G.; McGinnis, E.; Mitra, D.; Montgomery, R.; Pendola, A. The American Journal of Education:
Challenges and opportunities in translational science and the grey area of academic. Rev. Esp. Pedag. 2018,
76, 413–435. [CrossRef]
16. Magreñán, Á.A. Different anomalies in a Jarratt family of iterative root-finding methods. Appl. Math. Comput.
2014, 233, 29–38.
17. Magreñán, Á.A.; Argyros, I.K. A new tool to study real dynamics: The convergence plane. Appl. Math.
Comput. 2014, 248, 215–224. [CrossRef]
18. Magreñán, Á.A. On the local convergence and the dynamics of Chebyshev-Halley methods with six and
eight order of convergence. J. Comput. Appl. Math. 2016, 298, 236–251. [CrossRef]
19. Magreñán, Á.A.; Gutiérrez, J.M. Real dynamics for damped Newton’s method applied to cubic polynomials.
J. Comput. Appl. Math. 2015, 275, 527–538. [CrossRef]
20. Nakakoji, Y.; Wilson, R. First-Year Mathematics and Its Application to Science: Evidence of Transfer of
Learning to Physics and Engineering. Educ. Sci. 2018, 8, 8. [CrossRef]
21. Ortega, J.M.; Rheinboldt, W.C. Iterative Solution of Nonlinear Equations in Several Variables; Academic Press:
New York, NY, USA, 1970.
22. Rall, L.B. Convergence of Stirling’s method in Banach spaces. Aequ. Math. 1975, 12, 12–20. [CrossRef]
23. Rheinboldt, W.C. An adaptive continuation process for solving systems of nonlinear equations. Pol. Acad.
Sci. Banach. Cent. Publ. 1978, 3, 129–142. [CrossRef]
24. Traub, J.F. Iterative Methods for the Solution of Equations; Prentice-Hall Series in Automatic Computation:
Englewood Cliffs, NJ, USA, 1964.

c 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).