1 Introduction and preliminaries

The notions of α-ψ-contractive and α-admissible mappings were introduced by Samet et al. [1]. They proved some fixed point results for such mappings in complete metric spaces. These notions were generalized by Karapınar and Samet [2]. Asl et al. [3] extended these notions to multifunctions and introduced the notions of α -ψ-contractive and α -admissible mappings. Afterwards Ali and Kamran [4] further generalized the notion of α -ψ-contractive mappings and obtained some fixed point theorems for multivalued mappings. Some interesting extensions of results by Samet et al. [1] are available in [513]. Nadler initiated a fixed point theorem for multivalued mappings. Some extensions of Nadler’s result can also be found in [1431]. Mizoguchi and Takahashi [32] extended the Nadler fixed point theorem. Recently, Minak and Altun generalized Mizoguchi and Takahashi’s theorem by introducing a function α:X×X[0,). In this paper, we introduce the notion of α -Mizoguchi-Takahashi type contraction. By using this notion, we generalize some fixed point theorems presented by Minak and Altun [7], Kamran [26] and those contained therein.

We denote by CL(X) the class of all nonempty closed subsets of X and by CB(X) the class of all nonempty closed and bounded subsets of X. For ACL(X) or CB(X) and xX, d(x,A)=inf{d(x,a):aA}, and H is a generalized Hausdorff metric induced by d. Now we recollect some basic definitions and results for the sake of completeness.

If, for x 0 X, there exists a sequence { x n } in X such that x n T x n 1 , then O(T, x 0 )={ x 0 , x 1 , x 2 ,} is said to be an orbit of T:XCL(X) at x 0 . A mapping h:XR is said to be T-orbitally lower semicontinuous at ξX, if { x n } is a sequence in O(T, x 0 ) and x n ξ implies h(ξ)lim infh( x n ). The following definition is due to Asl et al. [3].

Definition 1.1 [3]

Let (X,d) be a metric space, α:X×X[0,) and T:XCL(X). Then T is α -admissible if for each x,yX with α(x,y)1 α (Tx,Ty)1, where α (Tx,Ty)=inf{α(a,b):aTx,bTy}.

Minak and Altun [7] generalized Mizoguchi and Takahashi’s theorem in the following way.

Theorem 1.2 [7]

Let (X,d) be a complete metric space, T:XCB(X) be a mapping satisfying

α (Tx,Ty)H(Tx,Ty)ϕ ( d ( x , y ) ) d(x,y)for each x,yX,

where ϕ:[0,)[0,1) such that lim sup r t + ϕ(r)<1 for each t[0,). Also assume that

  1. (i)

    T is α -admissible;

  2. (ii)

    there exists x 0 X with α( x 0 , x 1 )1 for some x 1 T x 0 ;

  3. (iii)
    1. (a)

      T is continuous,

    or

    1. (b)

      if { x n } is a sequence in X with x n x as n and α( x n , x n + 1 )1 for each nN{0}, then we have α( x n ,x)1 for each nN{0}.

Then T has a fixed point.

Kamran in [26] generalized Mizoguchi and Takahashi’s theorem in the following way.

Theorem 1.3 [26]

Let (X,d) be a complete metric space and T:XCL(X) be a mapping satisfying

d(y,Ty)ϕ ( d ( x , y ) ) d(x,y)for each xX and yTx,

where ϕ:[0,)[0,1) such that lim sup r t + ϕ(r)<1 for each t[0,). Then,

  1. (i)

    for each x 0 X, there exists an orbit { x n } of T and ξX such that lim n x n =ξ;

  2. (ii)

    ξ is a fixed point of T if and only if the function h(x):=d(x,Tx) is T-orbitally lower semicontinuous at ξ.

2 Main results

We begin this section with the following definition.

Definition 2.1 Let (X,d) be a metric space, T:XCL(X) is said to be an α -Mizoguchi-Takahashi type contraction if there exist two functions α:X×X[0,) and ϕ:[0,)[0,1) satisfying lim sup r t + ϕ(r)<1 for every t[0,) such that

α (Tx,Ty)d(y,Ty)ϕ ( d ( x , y ) ) d(x,y)for each xX and yTx.
(2.1)

Before moving toward our main results, we prove some lemmas.

Lemma 2.2 Let (X,d) be a metric space, { A k } be a sequence in CL(X), { x k } be a sequence in X such that x k A k 1 . Let ϕ:[0,)[0,1) be a function satisfying lim sup r t + ϕ(r)<1 for every t[0,). Suppose that {d( x k 1 , x k )} is a nonincreasing sequence such that

d( x k , A k )ϕ ( d ( x k 1 , x k ) ) d( x k 1 , x k ),
(2.2)
d( x k , x k + 1 )d( x k , A k )+ ϕ n k ( d ( x k 1 , x k ) ) ,
(2.3)

where n 1 < n 2 < , k, n k N. Then { x k } is a Cauchy sequence in X.

Proof The proof runs on the same lines as the proof of [[18], Lemma 3.2]. We include its details for completeness. Let d k :=d( x k 1 , x k ). Since d k is a nonincreasing sequence of nonnegative real numbers, therefore lim k d k =c0. By hypothesis, for t=c, we get lim sup t c + ϕ(t)<1. Therefore, there exists k 0 such that k k 0 implies that ϕ( d k )<h, where lim sup t c + ϕ(t)h<1. From (2.2) and (2.3), we have

d k + 1 ϕ ( d k ) d k + ϕ n k ( d k ) ϕ ( d k ) ϕ ( d k 1 ) d k 1 + ϕ ( d k ) ϕ n k 1 ( d k 1 ) + ϕ n k ( d k ) i = 1 k ϕ ( d i ) d 1 + m = 1 k 1 i = m + 1 k ϕ ( d i ) ϕ n m ( d m ) + ϕ n k ( d k ) i = 1 k ϕ ( d i ) d 1 + m = 1 k 1 i = max { k 0 , m + 1 } k ϕ ( d i ) ϕ n m ( d m ) + ϕ n k ( d k ) .
(2.4)

We have deleted some factors of ϕ from the product in (2.4) using the fact that ϕ<1. Let S denote the second term on the right-hand side of (2.4),

S ( k 0 1 ) h k k 0 + 1 m = 1 k 0 1 ϕ n m ( d m ) + m = k 0 k 1 h k m ϕ n m ( d m ) ( k 0 1 ) h k k 0 + 1 m = 1 k 0 1 ϕ n m ( d m ) + m = k 0 k 1 h k m + n m C h k + m = k 0 k 1 h k m + n m C h k + h k + n k 0 k 0 + h k + n k 0 1 ( k 0 1 ) + + h k + n k 1 ( k 1 ) C h k + m = k + n k 0 k 0 k + n k 1 ( k 1 ) h m = C h k + h k + n k 0 k 0 + 1 h k + n k 1 k + 2 1 h < C h k + h k h n k 0 k 0 + 1 1 h = C h k ,

where C is a generic positive constant. Now, it follows from (2.4) that

d k + 1 i = 1 k ϕ ( d i ) d 1 + C h k + ϕ n k ( d k ) < h k k 0 + 1 i = 1 k 0 1 ϕ ( d i ) d 1 + C h k + h n k < C h k + C h k + k = C h k ,

C again being a generic constant. Now, for k k 0 , mN,

d ( x k , x k + m ) i = k + 1 k + m d i < i = k + 1 k + m C h i 1 = C h k + 1 h k + m 1 h h k ,

which shows that { x k } is a Cauchy sequence in X. □

Lemma 2.3 Let (X,d) be a metric space, T:XCL(X) be an α -Mizoguchi-Takahashi type contraction. Let { x k } be an orbit of T at x 0 such that α (T x k 1 ,T x k )1 and

d( x k , x k + 1 )d( x k ,T x k )+ ϕ n k ( d ( x k 1 , x k ) ) ,
(2.5)

where x k T x k 1 , n 1 < n 2 < and k, n k N and {d( x k 1 , x k )} is a nonincreasing sequence. Then { x k } is a Cauchy sequence in X.

Proof Given that { x k } is an orbit of T at x 0 , i.e., x k T x k 1 for each kN, with α (T x k 1 ,T x k )1 for each kN, as T is an α -Mizoguchi-Takahashi type contraction. From (2.1), we have

d ( x k , T x k ) α ( T x k 1 , T x k ) d ( x k , T x k ) ϕ ( d ( x k 1 , x k ) ) d ( x k 1 , x k ) .

From (2.5), we have

d( x k , x k + 1 )d( x k ,T x k )+ ϕ n k ( d ( x k 1 , x k ) ) .

Since all the conditions of Lemma 2.2 are satisfied, { x k } is a Cauchy sequence in X. □

Theorem 2.4 Let (X,d) be a complete metric space, T:XCL(X) be an α -Mizoguchi-Takahashi type contraction and α -admissible. Suppose that there exist x 0 X and x 1 T x 0 such that α( x 0 , x 1 )1. Then,

  1. (i)

    there exists an orbit { x n } of T and x X such that lim x n = x ;

  2. (ii)

    x is a fixed point of T if and only if h(x)=d(x,Tx) is T-orbitally lower semicontinuous at x .

Proof By hypothesis, we have x 0 X and x 1 T x 0 with α( x 0 , x 1 )1. Thus, for x 1 T x 0 , we can choose a positive integer n 1 such that

ϕ n 1 ( d ( x 0 , x 1 ) ) [ 1 ϕ ( d ( x 0 , x 1 ) ) ] d( x 0 , x 1 ).
(2.6)

There exists x 2 T x 1 such that

d( x 1 , x 2 )d( x 1 ,T x 1 )+ ϕ n 1 ( d ( x 0 , x 1 ) ) .
(2.7)

As T is α -admissible, we have α (T x 0 ,T x 1 )1. From (2.6) and (2.7) it follows that

d ( x 1 , x 2 ) d ( x 1 , T x 1 ) + ϕ n 1 ( d ( x 0 , x 1 ) ) α ( T x 0 , T x 1 ) d ( x 1 , T x 1 ) + ϕ n 1 ( d ( x 0 , x 1 ) ) ϕ ( d ( x 0 , x 1 ) ) d ( x 0 , x 1 ) + [ 1 ϕ ( d ( x 0 , x 1 ) ) ] d ( x 0 , x 1 ) = d ( x 0 , x 1 ) .

Now we can choose a positive integer n 2 > n 1 such that

ϕ n 2 ( d ( x 1 , x 2 ) ) [ 1 ϕ ( d ( x 1 , x 2 ) ) ] d( x 1 , x 2 ).
(2.8)

There exists x 3 T x 2 such that

d( x 2 , x 3 )d( x 2 ,T x 2 )+ ϕ n 2 ( d ( x 1 , x 2 ) ) .
(2.9)

As T is α -admissible, then α( x 1 , x 2 ) α (T x 0 ,T x 1 )1 implies α (T x 1 ,T x 2 )1. Using (2.8) and (2.9) we have that

d ( x 2 , x 3 ) d ( x 2 , T x 2 ) + ϕ n 2 ( d ( x 1 , x 2 ) ) α ( T x 1 , T x 2 ) d ( x 2 , T x 2 ) + ϕ n 2 ( d ( x 1 , x 2 ) ) ϕ ( d ( x 1 , x 2 ) ) d ( x 1 , x 2 ) + [ 1 ϕ ( d ( x 1 , x 2 ) ) ] d ( x 1 , x 2 ) = d ( x 1 , x 2 ) .

By repeating this process for all kN, we can choose a positive integer n k such that

ϕ n k ( d ( x k 1 , x k ) ) [ 1 ϕ ( d ( x k 1 , x k ) ) ] d( x k 1 , x k ).
(2.10)

There exists x k T x k 1 such that

d( x k , x k + 1 )d( x k ,T x k )+ ϕ n k ( d ( x k 1 , x k ) ) .
(2.11)

Also, by α -admissibility of T, we have α (T x k 1 ,T x k )1 for each kN. From (2.10) and (2.11) it follows that

d ( x k , x k + 1 ) d ( x k , T x k ) + ϕ n k ( d ( x k 1 , x k ) ) α ( T x k 1 , T x k ) d ( x k , T x k ) + ϕ n k ( d ( x k 1 , x k ) ) ϕ ( d ( x k 1 , x k ) ) d ( x k 1 , x k ) + [ 1 ϕ ( d ( x k 1 , x k ) ) ] d ( x k 1 , x k ) = d ( x k 1 , x k ) ,

which implies that {d( x k , x k + 1 )} is a nonincreasing sequence of nonnegative real numbers. Thus, by Lemma 2.3, { x k } is a Cauchy sequence in X. Since X is complete, there exists x X such that x k x as k. Since x k T x k 1 , it follows from (2.1) that

d ( x k , T x k ) α ( T x k 1 , T x k ) d ( x k , T x k ) ϕ ( d ( x k 1 , x k ) ) d ( x k 1 , x k ) < d ( x k 1 , x k ) .

Letting k, in the above inequality, we have

lim k d( x k ,T x k )=0.
(2.12)

Suppose that h(x)=d(x,Tx) is T-orbitally lower semicontinuous at x , then

d ( x , T x ) =h ( x ) lim inf k h( x k )= lim inf k d( x k ,T x k )=0.

By the closedness of T it follows that x T x . Conversely, suppose that x is a fixed point of T, then h( x )=0 lim inf k h( x k ). □

Example 2.5 Let X={ 1 n :nN}{0}(1,) be endowed with the usual metric d. Define T:XCL(X) by

Tx={ { 0 } if  x = 0 , { 1 n + 2 , 1 n + 3 } if  x = 1 n : 1 n 6 , { 1 n , 0 } if  x = 1 n : n > 6 , [ 2 x , ) if  x > 1 ,

and α:X×X[0,) by

α(x,y)={ 1 if  x , y { 1 n : n N } { 0 } , 0 otherwise .

Define ϕ:[0,)[0,1) by

ϕ(t)={ 4 5 if  0 t 1 6 , 1 2 if  t > 1 6 .

One can check that for each xX and yTx, we have

α (Tx,Ty)d(y,Ty)ϕ ( d ( x , y ) ) d(x,y).

Also, T is α -admissible and for x 0 =1 we have x 1 = 1 3 T x 0 with α( x 0 , x 1 )=1. Moreover, all the other conditions of Theorem 2.4 are satisfied. Therefore T has a fixed point. Note that Theorem 5 of Minak and Altun [7] is not applicable here; see, for example, x= 1 7 and y= 1 8 . Further Theorem 2.1 of Kamran [26] is also not applicable; see, for example, x=2 and y=4Tx.

The proofs of the following theorems run on the same lines as the proof of Theorem 2.4.

Theorem 2.6 Let (X,d) be a complete metric space, T:XCL(X) be an α -admissible mapping such that

α (y,Ty)d(y,Ty)ϕ ( d ( x , y ) ) d(x,y)for each xX and yTx,
(2.13)

where ϕ:[0,)[0,1) satisfying lim sup r t + ϕ(r)<1 for every t[0,). Suppose that there exist x 0 X and x 1 T x 0 such that α( x 0 , x 1 )1. Then,

  1. (i)

    there exists an orbit { x n } of T and x X such that lim x n = x ;

  2. (ii)

    x is a fixed point of T if and only if h(x)=d(x,Tx) is T-orbitally lower semicontinuous at x .

Theorem 2.7 Let (X,d) be a complete metric space, T:XCL(X) be an α -admissible mapping such that

α(x,y)d(y,Ty)ϕ ( d ( x , y ) ) d(x,y)for each xX and yTx,
(2.14)

where ϕ:[0,)[0,1) satisfying lim sup r t + ϕ(r)<1 for every t[0,). Suppose that there exist x 0 X and x 1 T x 0 such that α( x 0 , x 1 )1. Then,

  1. (i)

    there exists an orbit { x n } of T and x X such that lim x n = x ;

  2. (ii)

    x is a fixed point of T if and only if h(x)=d(x,Tx) is T-orbitally lower semicontinuous at x .

Corollary 2.8 [26]

Let (X,d) be a complete metric space and T:XCL(X) be a mapping satisfying

d(y,Ty)ϕ ( d ( x , y ) ) d(x,y)for each xX and yTx,

where ϕ:[0,)[0,1) such that lim sup r t + ϕ(r)<1 for each t[0,). Then,

  1. (i)

    for each x 0 X, there exists an orbit { x n } of T and ξX such that lim n x n =ξ;

  2. (ii)

    ξ is a fixed point of T if and only if the function h(x):=d(x,Tx) is T-orbitally lower semicontinuous at ξ.

Proof Define α:X×X[0,) by α(x,y)=1 for each x,yX. Then the proof follows from Theorem 2.4 as well as from Theorem 2.6, and from Theorem 2.7. □

3 Application

From Definition 2.1, we get the following definition by considering only those xX and yTx for which we have α (Tx,Ty)1.

Definition 3.1 Let (X,d) be a metric space, T:XCL(X) is said to be a modified α -Mizoguchi-Takahashi type contraction if there exist two functions α:X×X[0,) and ϕ:[0,)[0,1) satisfying lim sup r t + ϕ(r)<1 for every t[0,) such that for each xX and yTx,

α (Tx,Ty)1d(y,Ty)ϕ ( d ( x , y ) ) d(x,y).
(3.1)

Lemma 3.2 Let (X,d) be a metric space, T:XCL(X) be a modified α -Mizoguchi-Takahashi contraction. Let { x k } be an orbit of T at x 0 such that α (T x k 1 ,T x k )1 and

d( x k , x k + 1 )d( x k ,T x k )+ ϕ n k ( d ( x k 1 , x k ) ) ,
(3.2)

where x k T x k 1 , n 1 < n 2 < and k, n k N and {d( x k 1 , x k )} is a nonincreasing sequence. Then { x k } is a Cauchy sequence in X.

Proof Given that { x k } is an orbit of T at x 0 , i.e., x k T x k 1 for each kN, with α (T x k 1 ,T x k )1 for each kN, as T is a modified α -Mizoguchi-Takahashi contraction. From (3.1), we have

d( x k ,T x k )ϕ ( d ( x k 1 , x k ) ) d( x k 1 , x k ).

From (3.2), we have

d( x k , x k + 1 )d( x k ,T x k )+ ϕ n k ( d ( x k 1 , x k ) ) .

Since all the conditions of Lemma 2.2 are satisfied, { x k } is a Cauchy sequence in X. □

Working on the same lines as the proof of Theorem 2.4 is done, one may obtain the proof of the following result.

Theorem 3.3 Let (X,d) be a complete metric space, T:XCL(X) be a modified α -Mizoguchi-Takahashi contraction and α -admissible. Suppose that there exist x 0 X and x 1 T x 0 such that α( x 0 , x 1 )1. Then,

  1. (i)

    there exists an orbit { x n } of T and x X such that lim x n = x ;

  2. (ii)

    x is a fixed point of T if and only if h(x)=d(x,Tx) is T-orbitally lower semicontinuous at x .