In this paper, we introduce a new hybrid relaxed iterative algorithm with two half-spaces to solve the fixed-point problem and split-feasibility problem involving demicontractive mappings. The strong convergence of the iterative sequence produced by our algorithm is proved under certain weak conditions. We give several numerical experiments to demonstrate the efficiency of the proposed iterative method in comparison with previous algorithms.
Citation: Yuanheng Wang, Bin Huang, Bingnan Jiang, Tiantian Xu, Ke Wang. A general hybrid relaxed CQ algorithm for solving the fixed-point problem and split-feasibility problem[J]. AIMS Mathematics, 2023, 8(10): 24310-24330. doi: 10.3934/math.20231239
Meiying Wang, Luoyi Shi .
A new self-adaptive inertial algorithm with $ W $-mapping for solving split feasibility problem in Banach spaces. AIMS Mathematics, 2022, 7(10): 18767-18783.
doi: 10.3934/math.20221032
[3]
Jamilu Abubakar, Poom Kumam, Jitsupa Deepho .
Multistep hybrid viscosity method for split monotone variational inclusion and fixed point problems in Hilbert spaces. AIMS Mathematics, 2020, 5(6): 5969-5992.
doi: 10.3934/math.2020382
[4]
Lu-Chuan Ceng, Shih-Hsin Chen, Yeong-Cheng Liou, Tzu-Chien Yin .
Modified inertial subgradient extragradient algorithms for generalized equilibria systems with constraints of variational inequalities and fixed points. AIMS Mathematics, 2024, 9(6): 13819-13842.
doi: 10.3934/math.2024672
[5]
Meiying Wang, Luoyi Shi, Cuijuan Guo .
An inertial iterative method for solving split equality problem in Banach spaces. AIMS Mathematics, 2022, 7(10): 17628-17646.
doi: 10.3934/math.2022971
[6]
Yajun Xie, Changfeng Ma .
The hybird methods of projection-splitting for solving tensor split feasibility problem. AIMS Mathematics, 2023, 8(9): 20597-20611.
doi: 10.3934/math.20231050
[7]
Mohammad Dilshad, Aysha Khan, Mohammad Akram .
Splitting type viscosity methods for inclusion and fixed point problems on Hadamard manifolds. AIMS Mathematics, 2021, 6(5): 5205-5221.
doi: 10.3934/math.2021309
[8]
Yasir Arfat, Poom Kumam, Supak Phiangsungnoen, Muhammad Aqeel Ahmad Khan, Hafiz Fukhar-ud-din .
An inertially constructed projection based hybrid algorithm for fixed point and split null point problems. AIMS Mathematics, 2023, 8(3): 6590-6608.
doi: 10.3934/math.2023333
[9]
Junaid Ahmad, Kifayat Ullah, Reny George .
Numerical algorithms for solutions of nonlinear problems in some distance spaces. AIMS Mathematics, 2023, 8(4): 8460-8477.
doi: 10.3934/math.2023426
[10]
Charu Batra, Renu Chugh, Mohammad Sajid, Nishu Gupta, Rajeev Kumar .
Generalized viscosity approximation method for solving split generalized mixed equilibrium problem with application to compressed sensing. AIMS Mathematics, 2024, 9(1): 1718-1754.
doi: 10.3934/math.2024084
Abstract
In this paper, we introduce a new hybrid relaxed iterative algorithm with two half-spaces to solve the fixed-point problem and split-feasibility problem involving demicontractive mappings. The strong convergence of the iterative sequence produced by our algorithm is proved under certain weak conditions. We give several numerical experiments to demonstrate the efficiency of the proposed iterative method in comparison with previous algorithms.
1.
Introduction
The split feasibility problem (SFP) in Euclidean spaces was introduced by Censor and Elfving[1] in 1994. It is depicted as finding a point x such that
x∈C,Ax∈Q,
(1.1)
where C⊂H1 and Q⊂H2 are nonempty closed convex sets; A is a bounded linear operator from a Hilbert space H1 onto a Hilbert space H2 with A≠0. Denote the set of solutions for (1.1) by Δ. The SFP plays an important role in the study of medical image reconstruction, signal processing, etc. Many scholars regard the SFP and its generalizations, such as the multiple-set SFP and the split common-fixed-point problem as their research direction (see [2,3,4,5,6]).
It is known that Byrne [7,8] proposed the famous CQ algorithm (CQA) for solving the problem (1.1). The proposed method is given as follows:
xn+1=PC(xn−τnA∗(I−PQ)Axn),∀n≥1,
(1.2)
where τn∈(0,2‖A‖2) is the step size, A∗ is the adjoint operator of A, PC is the projection onto C and PQ is the projection onto Q. Regrettably, there are two drawbacks of applying (1.2). On one hand, estimating the operator norm can often be challenging. On the other hand, computing projections onto both C and Q can be very difficult.
To overcome the first drawback, López et al. [9] proposed a novel approach for selecting the step size, denoted by τn, which is defined as follows:
τn=ρnf(xn)‖∇f(xn)‖2,
where f(xn):=12‖(I−PQ)Axn‖2, ∇f(xn)=A∗(I−PQ)Axn, ρn∈(0,4) and infn∈Nρn(4−ρn)>0.
To overcome the second drawback, Fukushima [10] introduced the level sets C and Q:
C={x∈H1:ϕ(x)≤0},Q={y∈H2:φ(y)≤0},
(1.3)
where ϕ:H1→R and φ:H2→R are convex, subdifferential and weakly lower semi-continuous functions. Motivated by Fukushima's method, Yang [11] introduced a new relaxed CQA by replacing PC with PCn and PQ with PQn. The algorithm has the form
xn+1=PCn(xn−αnA∗(I−PQn)Axn),∀n≥1,
where αn∈(0,2‖AA∗‖), A∗ is the adjoint operator of A and Cn and Qn are defined as follows:
Cn={x∈H1:ϕ(xn)≤⟨ϑn,xn−x⟩},ϑn∈∂ϕ(xn),
and
Qn={y∈H2:φ(Axn)≤⟨χn,Axn−y⟩},χn∈∂φ(Axn),
where ∂ϕ and ∂φ are bounded operators.
Qin and Wang [12] introduced (1.4) to find a common solution of the SFP (1.1) and fixed-point problem in 2019. Their algorithm has the form
where g is a k-contraction from C onto C, S is nonexpansive from C onto C and Fix(S):={x∈H:Sx=x}. {αn}, {βn}, {γn}, {δn} and {τn} are included in (0,1); they satisfy the following:
Then, there is a point q of Δ∩Fix(S) with xn→q. Furthermore, it satisfies (1.5):
⟨z′−q,g(q)−q⟩≤0,∀z′∈Fix(S)∩Δ.
(1.5)
Motivated by López et al. [9], Wang et al. [13] improved the selection of step size in (1.4). In order to accelerate the convergence rate, they introduced (1.6) to solve the SFP (1.1) and fixed-point problem. Given that x0,x1∈H, their algorithm has the form
where μ≥0, τn=ρnf(wn)‖∇f(wn)‖2, 0<ρn<4, f(wn):=12‖(I−PQ)Awn‖2 and ∇f(wn)=A∗(I−PQ)Awn. Assume that S:H1→H1 is quasi-nonexpansive, I−S is demiclosed at zero and g:H1→H1 is a k-contraction. {αn}, {βn}, {γn} and {δn} are included in [0,1], which satisfy the following conditions:
(i) ∑∞n=1αn=∞, limn→∞αn=0;
(ii) lim supn→∞βn<1;
(iii) αn+βn+γn=1;
(iv) 0<lim infn→∞δn≤lim supn→∞δn<1;
(v) εn>0, limn→∞εnαn=0.
Then, there is a point q of Fix(S)∩Δ with xn→q. Furthermore, it satisfies (1.7):
⟨z′−q,g(q)−q⟩≤0,∀z′∈Fix(S)∩Δ.
(1.7)
In addition, Tian [14] introduced an iterative form:
xn+1=αnλg(xn)+(I−ιαnB)Sxn,∀n∈N.
Suppose that Fix(S)≠∅, where Fix(S):={x∈H:Sx=x}. Then, there is a point q in Fix(S) with xn→q. Furthermore, it satisfies the following:
⟨(ιB−λg)q,z′−q⟩≥0,∀z′∈Fix(S),
where B is ¯λ-strongly monotone and M-Lipschitz continuous from H onto H with ¯λ,M>0, S is nonexpansive from H onto H and g is a k-contraction from H onto H with 0<k<1. Let ι,λ∈R and {αn}⊂(0,1) satisfy the following:
(i) 0<ι<2¯λM2;
(ii) 0<λ<ι¯λ−M2ι2k;
(iii) ∑∞n=1|αn+1−αn|<∞, ∑∞n=1αn=∞, limn→∞αn=0.
Motivated by the algorithms developed by Qin and Wang [12], Wang et al. [13], Tian [14] and Kwari et al. [15], we propose a new half-space relaxation projection method for solving the SFP and fixed-point problem of demicontractive mappings by expanding the scope of operator S, modifying the iteration format and optimizing the selection of step sizes. Our result in this article extends and improves many recent correlative results of other authors. Particularly, our method extends and improves the methods in some papers of other authors in the following ways: (1) We have considered and studied a new modified half-space hybrid method to solve the fixed-point problem and SFP of nonlinear operators at the same time. And, our method can be used more widely; (2) We extend the nonexpansive mapping and the quasi-nonexpansive mapping to the demicontractive mapping, which expands the scope of the study; (3) The inertia is added to accelerate the convergence rate further; (4) The selection of the step size is a multistep and self-adaptive process, so it no longer depends on the operator norm; (5) The projection onto a half-space greatly facilitates the calculation, and under some weaker conditional assumptions, the iterative sequence generated by our new algorithm converges strongly to the common solution of the two problems. At last, we present some numerical experiments to show the effectiveness and feasibility of our new iterative method.
2.
Preliminaries
In this paper, let H be a real Hilbert space; the inner product be denoted by ⟨⋅,⋅⟩ and the norm be denoted by ‖⋅‖. We use the symbol ⇀ to indicate weak convergence and → to indicate strong convergence.
Definition 2.1. A self-mapping P is said to be
(i) nonexpansive if
‖Px−Py‖≤‖x−y‖,∀x,y∈H;
(ii) quasi-nonexpansive if Fix(P)≠∅ such that
‖Px−y‖≤‖x−y‖,∀x∈H,y∈Fix(P);
(iii)α -demicontractive (0≤α<1) if Fix(P)≠∅ such that
‖Px−y‖2≤α‖(I−P)x‖2+‖x−y‖2,∀x∈H,y∈Fix(P);
(iv) firmly nonexpansive if
‖Px−Py‖2≤⟨Px−Py,x−y⟩,∀x,y∈H;
(v)t-contractive (0≤t<1) if
‖Px−Py‖≤t‖x−y‖,∀x,y∈H;
(vi)M-Lipschitz continuous (M>0) if
‖Px−Py‖≤M‖x−y‖,∀x,y∈H;
(vii)β-strongly monotone (β≥0) if
⟨Px−Py,x−y⟩≥β‖x−y‖2,∀x,y∈H.
When Fix(P)≠∅, we obtain (i)⇒(ii)⇒(iii).
Definition 2.2. When the function f:H→R is convex and subdifferentiable, an element d∈H is called a subgradient of f(x0) if
f(y)≥f(x0)+⟨d,y−x0⟩,∀y∈H.
The set of subgradients of f at x0 is denoted by ∂f(x0).
Definition 2.3. The function f:H→R is called weakly lower semi-continuous if xn⇀x0 implies that
f(x0)≤lim infn→∞f(xn).
Lemma 2.1 ([16,17,18]). Let D⊂H be a nonempty, convex and closed set. For any x∈H, we have
Lemma 2.3 ([14]). Assume that B is an M-Lipschitz continuous and ¯λ-strongly monotone operator with ¯λ,M>0. For μ,ν>0, they satisfy the following conditions:
0<μ<2¯λM2,ν=¯λ−M2μ2.
Let {γn}⊂(0,1) and limn→∞γn=0. If n is a sufficiently large number, we have
Lemma 2.5 ([8]). Let f(z)=12‖(I−PQ)Az‖2. We can know that ∇f is M-Lipschitz continuous with M=‖A‖2.
Lemma 2.6 ([22]). Let {an} be a sequence of nonnegative numbers such that
an+1≤(1−Θn)an+ΘnΥn,n≥n0,an+1≤an−Πn+Γn,n≥n0,
where n0 is a sufficiently large integer, the sequence {Θn} is contained in (0,1), the real sequence {Πn} is nonnegative {Υn} and {Γn} are two sequences which are included in R. If the following conditions hold
(i)∑∞n=0Θn=∞; (ii)limn→∞Γn=0;
(iii)limi→∞Πni=0⇒lim supi→∞Υni≤0, ∀{ni}⊆{n}, then limn→∞an=0.
3.
Results
In this section, C and Q are defined by (1.3). We give some conditions for the convergence analysis of our Algorithm 1.
(C1)∑∞n=1αn=∞, limn→∞αn=0, 0<αn<1;
(C2)βn∈[0,1], lim supn→∞βn<1;
(C3)εn>0, limn→∞εnαn=0;
(C4)0<lim infn→∞δn≤lim supn→∞δn<1;
(C5)σn∈(0,1−α], lim infn→∞σn>0;
(C6)infn∈Nρn(4−ρn)>0, {ρn}⊂(0,4).
Algorithm 1
Initialization: Apply x0,x1∈H1, μ≥0, τ>0, 0<ι<2¯λL2, ν=¯λ−L2ι2, λ>0 such that 0≤kλ<ν.
Theorem 3.1.Let S:H1→H1 be α-demicontractive, I−S be demiclosed at zero and g:H1→H1 be k-contractive. Moreover, let operator B be L-Lipschitz continuous and ¯λ-strongly monotone. Assume that {αn}, {βn}, {δn}, {σn}, {ρn} are sequences satisfying (C1)−(C6) and Fix(S)∩Δ≠∅. Then, the {xn} generated by Algorithm 1 converges strongly to x∗∈Fix(S)∩Δ. Furthermore, it is the unique solution of (3.1):
⟨(B−λg)x∗,z′−x∗⟩≥0,∀z′∈Fix(S)∩Δ.
(3.1)
Proof. Let r∈Δ∩Fix(S). Since C⊂Cn and Q⊂Qn, we have that r∈Cn, Ar∈Qn and r∈Fix(S). By combining Lemma 2.1 with Lemma 2.2, we get
It is easy for us to know that limn→∞Θn=0, ∑∞n=1Θn=∞ andlimn→∞Γn=0. If we prove that lim supi→∞Υni≤0 when limi→∞Πni=0 for any subsequence {ni}⊂{n}, we can get that limn→∞‖xn−x∗‖=0 by Lemma 2.6.
Assume that
limi→∞Πni=0.
Using the conditions of {βn}, {δn} and {σn}, we have
It is easy for us to know that ωw(dni)⊂Fix(S) by combining (3.18) and the case that I−S is demiclosed at zero. Now, we choose a subsequence {dnij} of {dni} to satisfy
Without loss of generality, we suppose that dnij⇀z′. It is easy to get that ynij⇀z′ and xnij⇀z′ by using (3.21) and (3.23). Now, we show that z′∈C. We know that ynij∈Cnijby the definition of ynij; this implies that
ϕ(xnij)≤⟨ϑnij,xnij−ynij⟩,
(3.24)
where ϑnij∈∂ϕ(xnij). From (3.23) and the boundedness of ∂ϕ, we have
ϕ(xnij)≤‖ϑnij‖‖xnij−ynij‖→0
(3.25)
as n→∞. Because of the weakly lower semi-continuous nature of ϕ, xnij⇀z′ and (3.25), the following holds:
ϕ(z′)≤lim infj→∞ϕ(xnij)≤0.
Hence z′∈C.
Next, we show that Az′∈Q. Since PQnij(Axnij)∈Qnij, we have
φ(Axnij)≤⟨χnij,Axnij−PQnij(Axnij)⟩,
(3.26)
where χnij∈∂φ(Axnij). Given that limi→∞‖(I−PQni)Adni‖=0 and (3.22), we get that limi→∞‖(I−PQni)Axni‖=0. By (3.26) and the boundedness of ∂φ, we have
φ(Axnij)≤‖χnij‖‖Axnij−PQnij(Axnij)‖→0
(3.27)
as n→∞. In view of xnij⇀z′, Axnij⇀Az′, the weakly lower semi-continuous nature of φ and (3.27), we can get
φ(Az′)≤lim infj→∞φ(Axnij)≤0.
Hence Az′∈Q.
From the above proof process, we can derive that z′∈Fix(S)∩Δ. Now, using (3.21), we get
Now, we present a comparison of Algorithm 1 with other algorithms by describing two numerical experiments. All of the programs were written in Matlab 9.5 and performed on a desktop PC with AMD Ryzen 7 4800U with Radeon Graphics 1.80 GHz, RAM 16.0GB.
Example 4.1. We assume that H1=H2=C=R4 and Q={b}, where b∈R4; then, we have that Δ={x∈R4:Ax=b}. Considering the system of linear equations Ax=b, let
A=(112323523−1−1−2352−2),b=(51−6−15).
By calculation, we can find that Ax=b has a unique solution x∗=(−1,−2,1,2)T. Let
Sx=(14140001414000141400034)x+(−14−741412),∀x∈R4.
By calculation, we can find that the mapping S is nonexpansive and x∗=(−1,−2,1,2)T∈Fix(S).
In Algorithm 1 and scheme (1.4), we let αn=110n, βn=0.2, δn=0.5, x1=(1,2,3,4)T and g=0.2I; in scheme (1.4), we let τn=74‖A‖2; in Algorithm 1, we let ϕ(x)=0, ∀x∈R4, φ(y)=‖y−b‖, ∀y∈R4, B=I, λ=1, εn=1n2, μ=1.5, σn=1, ρn=3.5, τ=1 and x0=(1,2,3,4)T. The results of numercial experiments are revealed in Table 1 and Figure 1.
Table 1.
Numerical results of Algorithm 1 for Example 4.1.
From Table 1, we can find that xn is closer to the exact solution x∗ with an increase of the number of iterations. We can also see that errors are closer to 0. Therefore, it can be concluded that Algorithm 1 is feasible. From Figure 1, we can find that the number of iterations for Algorithm 1 is less than that for scheme (1.4).
Example 4.2. Assume that H1=R, H2=R3, C=[−2,6], Q={(ya,yb,yc)T:|ya|+|yb|+|yc|≤3},
A=(2−34),
Sx=x2sinx,∀x∈R.
We can see that S is not nonexpansive but quasi-nonexpansive [23]. By calculation, we can find that Fix(S)={0}. It is easy to obtain that 0∈Δ. We denote x∗=0.
In Algorithm 1 and scheme (1.6), we let x0=x1=1, εn=1n2, μ=0.9, βn=0.5, ρn=3.5, δn=0.4, g(x)=13sinx and αn=18np, where 0<p≤1; in scheme (1.6), we used the function quadprog to compute the projection over Q by using Matlab 9.5 Optimization Toolbox; In Algorithm 1, we let ϕ(x)=(x−6)(x+2), ∀x∈R, φ(y)=|ya|+|yb|+|yc|−3, ∀y=(ya,yb,yc)T∈R3, B=I, λ=1, σn=1 and τ=1. We used ‖xn−x∗‖≤10−8 as the stopping criterion. The results of numerical experiments for different values of p are shown in Table 2. The convergence behavior for p=0.5 is shown in Figures 2 and 3.
Table 2.
Numerical results of Algorithm 1 and scheme (1.6) for Example 4.2.
In this work, we developed a method for solving the SFP and the fixed-point problem involving demicontractive mapping. In Algorithm 1, we have expanded the nonexpansive mapping to the demicontractive mapping, selected a new step size and added the inertial method based on scheme (1.4). Our Algorithm 1 exhibits better convergence behavior than scheme (1.4). Additionally, we have extended the quasi-nonexpansive mapping to the demicontractive mapping, applied projection onto a half-space and selected a new step size in our Algorithm 1 based on scheme (1.6). Our Algorithm 1 also demonstrated superior convergence behavior compared to scheme (1.6). From the above, our algorithm exhibits a superior convergence rate and a wider range for the operator, thus making it applicable in a broader range of scenarios. Furthermore, we have presented two numerical results that demonstrate the feasibility and effectiveness of our method. In the following work, we intend to extend the demicontractive mapping to the quasi-pseudocontractive mapping and extend the SFP to the multiple-set split feasibility problem.
Use of AI tools declaration
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Grant no. 12171435).
Conflict of interest
The authors declare that they have no competing interests.
References
[1]
Y. Censor, T. Elfving, A multiprojection algorithm using Bregman projections in a product space, Numer. Algor., 8 (1994), 221–239. https://doi.org/10.1007/BF02142692 doi: 10.1007/BF02142692
[2]
Y. Yao, M. Postolache, X. Qin, J. Yao, Iterative algorithms for the proximal split feasibility problem, U. P. B. Sci. Bull. A, 80 (2018), 37–44.
[3]
A. E. Al-Mazrooei, A. Latif, X. Qin, J. C. Yao, Fixed point algorithms for split feasibility problems, Fixed Point Theor., 20 (2019), 245–254. https://doi.org/10.24193/fpt-ro.2019.1.16 doi: 10.24193/fpt-ro.2019.1.16
[4]
Q. Dong, S. He, Y. Zhao, On global convergence rate of two acceleration projection algorithms for solving the multiple-sets split feasibility problem, Filomat, 30 (2016), 3243–3252. https://doi.org/10.2298/FIL1612243D doi: 10.2298/FIL1612243D
[5]
J. Zhao, D. Hou, A self-adaptive iterative algorithm for the split common fixed problems, Numer Algor., 82 (2019), 1047–1063. https://doi.org/10.1007/s11075-018-0640-x doi: 10.1007/s11075-018-0640-x
[6]
D. Hou, J. Zhao, X. Wang, Weak convergence of a primal-dual algorithm for split common fixed-point problems in Hilbert spaces, J. Appl. Numer. Optim., 2 (2020), 187–197.
[7]
C. Byrne, Iterative oblique projection onto convex set and the split feasibility problem, Inverse Probl., 18 (2002), 441–453. https://doi.org/10.1088/0266-5611/18/2/310 doi: 10.1088/0266-5611/18/2/310
[8]
C. Byrne, A unified treatment of some iterative algorithms in signal processing and image reconstruction, Inverse Probl., 20 (2004), 103–120. https://doi.org/10.1088/0266-5611/20/1/006 doi: 10.1088/0266-5611/20/1/006
[9]
G. López, V. Martín-Márquez, F. Wang, H. K. Xu, Solving the split feasibility problem without prior knowledge of matrix norms, Inverse Probl., 28 (2012), 085004. https://doi.org/10.1088/0266-5611/28/8/085004 doi: 10.1088/0266-5611/28/8/085004
[10]
M. Fukushima, A relaxed projection method for variational inequalities, Math. Program., 35 (1986), 58–70. https://doi.org/10.1007/BF01589441 doi: 10.1007/BF01589441
[11]
Q. Yang, The relaxed CQ algorithm for solving the split feasibility problem, Inverse Probl., 20 (2004), 1261–1266. https://doi.org/10.1088/0266-5611/20/4/014 doi: 10.1088/0266-5611/20/4/014
[12]
X. Qin, L. Wang, A fixed point method for solving a split feasibility problem in Hilbert spaces, RACSAM, 13 (2019), 215–325. https://doi.org/10.1007/s13398-017-0476-6 doi: 10.1007/s13398-017-0476-6
[13]
Y. H. Wang, T. T. Xu, J. C. Yao, B. N. Jiang, Self-adaptive method and inertial modification for solving the split feasibility problem and fixed-point problem of quasi-nonexpansive mapping, Mathematics, 10 (2022), 1612. https://doi.org/10.3390/math10091612 doi: 10.3390/math10091612
[14]
M. Tian, A general iterative algorithm for nonexpansive mapping in Hilbert spaces, Nonlinear Anal., 73 (2010), 689–694. https://doi.org/10.1016/j.na.2010.03.058 doi: 10.1016/j.na.2010.03.058
H. H. Bauschke, P. L. Combettes, Convex analysis and monotone operator theory in Hilbert space, New York: Springer, 2011.
[17]
S. Kesornprom, N. Pholasa, P. Cholamjiak, On the convergence analysis of the gradient-CQ algorithms for the split feasibility problem, Numer. Alogr., 84 (2020), 997–1017. https://doi.org/10.1007/s11075-019-00790-y doi: 10.1007/s11075-019-00790-y
[18]
S. Suantai, S. Kesornprom, N. Pholasa, Y. J. Cho, P. Cholamjiak, A relaxed projections method using a new linesearch for the split feasibility problem, AIMS Mathematics, 6 (2021), 2690–2703.
[19]
Y. H. Wang, M. Y. Yuan, B. N. Jiang, Multi-step inertial hybrid and shrinking Tseng's algorithm with Meir-Keeler contractions for variational inclusion problems, Mathematics, 9 (2021), 1548. https://doi.org/10.3390/math9131548 doi: 10.3390/math9131548
[20]
W. L. Sun, G. Lu, Y. F. Jin, C. Park, Self-adaptive algorithms for the split problem of the quasi-pseudocontractive operators in Hilbert spaces, AIMS Mathematics, 7 (2022), 8715–8732. https://doi.org/10.3934/math.2022487 doi: 10.3934/math.2022487
[21]
D. V. Thong, Viscosity approximation methods for solving fixed point problems and split common fixed point problems, J. Fixed Point Theory Appl., 19 (2017), 1481–1499. https://doi.org/10.1007/s11784-016-0323-y doi: 10.1007/s11784-016-0323-y
[22]
S. He, C. Yang, Solving the variational inequality problem defined on intersection of finite level sets, Abstr. Appl. Anal., 2013 (2013), 942315. https://doi.org/10.1155/2013/942315 doi: 10.1155/2013/942315
[23]
M. Tian, M. Tong, Self-adaptive subgradient extragradient method with inertial modification for solving monotone variational inequality problems and quasi-nonexpansive fixed point problems, J. Inequal. Appl., 2019 (2019), 7. https://doi.org/10.1186/s13660-019-1958-1 doi: 10.1186/s13660-019-1958-1
Yuanheng Wang, Bin Huang, Bingnan Jiang, Tiantian Xu, Ke Wang. A general hybrid relaxed CQ algorithm for solving the fixed-point problem and split-feasibility problem[J]. AIMS Mathematics, 2023, 8(10): 24310-24330. doi: 10.3934/math.20231239
Yuanheng Wang, Bin Huang, Bingnan Jiang, Tiantian Xu, Ke Wang. A general hybrid relaxed CQ algorithm for solving the fixed-point problem and split-feasibility problem[J]. AIMS Mathematics, 2023, 8(10): 24310-24330. doi: 10.3934/math.20231239