Loading [MathJax]/jax/output/SVG/jax.js
Research article

Modeling of 3 SAT discrete Hopfield neural network optimization using genetic algorithm optimized K-modes clustering

  • Received: 07 August 2024 Revised: 05 September 2024 Accepted: 18 September 2024 Published: 27 September 2024
  • MSC : 03B52, 68T27, 68N17, 68W99

  • The discrete Hopfield neural network 3-satisfiability (DHNN-3SAT) model represents an innovative application of deep learning techniques to the Boolean SAT problem. Existing research indicated that the DHNN-3SAT model demonstrated significant advantages in handling 3SAT problem instances of varying scales and complexities. Compared to traditional heuristic algorithms, this model converged to local minima more rapidly and exhibited enhanced exploration capabilities within the global search space. However, the model faced several challenges and limitations. As constraints in SAT problems dynamically increased, decreased, or changed, and as problem scales expanded, the model's computational complexity and storage requirements may increase dramatically, leading to reduced performance in handling large-scale SAT problems. To address these challenges, this paper first introduced a method for designing network synaptic weights based on fundamental logical clauses. This method effectively utilized the synaptic weight information from the original SAT problem within the DHNN network, thereby significantly reducing redundant computations. Concrete examples illustrated the design process of network synaptic weights when constraints were added, removed, or updated, offering new approaches for managing the evolving constraints in SAT problems. Subsequently, the paper presented a DHNN-3SAT model optimized by genetic algorithms combined with K-modes clustering. This model employed genetic algorithm-optimized K-modes clustering to effectively cluster the initial space, significantly reducing the search space. This approach minimized the likelihood of redundant searches and reduced the risk of getting trapped in local minima, thus improving search efficiency. Experimental tests on benchmark datasets showed that the proposed model outperformed traditional DHNN-3SAT models, DHNN-3SAT models combined with genetic algorithms, and DHNN-3SAT models combined with imperialist competitive algorithms across four evaluation metrics. This study not only broadened the application of DHNN in solving 3SAT problems but also provided valuable insights and guidance for future research.

    Citation: Xiaojun Xie, Saratha Sathasivam, Hong Ma. Modeling of 3 SAT discrete Hopfield neural network optimization using genetic algorithm optimized K-modes clustering[J]. AIMS Mathematics, 2024, 9(10): 28100-28129. doi: 10.3934/math.20241363

    Related Papers:

    [1] K. R. Karthikeyan, G. Murugusundaramoorthy, N. E. Cho . Some inequalities on Bazilevič class of functions involving quasi-subordination. AIMS Mathematics, 2021, 6(7): 7111-7124. doi: 10.3934/math.2021417
    [2] Muajebah Hidan, Abbas Kareem Wanas, Faiz Chaseb Khudher, Gangadharan Murugusundaramoorthy, Mohamed Abdalla . Coefficient bounds for certain families of bi-Bazilevič and bi-Ozaki-close-to-convex functions. AIMS Mathematics, 2024, 9(4): 8134-8147. doi: 10.3934/math.2024395
    [3] Ala Amourah, B. A. Frasin, G. Murugusundaramoorthy, Tariq Al-Hawary . Bi-Bazilevič functions of order ϑ+iδ associated with (p,q) Lucas polynomials. AIMS Mathematics, 2021, 6(5): 4296-4305. doi: 10.3934/math.2021254
    [4] Mohsan Raza, Khalida Inayat Noor . Subclass of Bazilevič functions of complex order. AIMS Mathematics, 2020, 5(3): 2448-2460. doi: 10.3934/math.2020162
    [5] Bakhtiar Ahmad, Muhammad Ghaffar Khan, Basem Aref Frasin, Mohamed Kamal Aouf, Thabet Abdeljawad, Wali Khan Mashwani, Muhammad Arif . On q-analogue of meromorphic multivalent functions in lemniscate of Bernoulli domain. AIMS Mathematics, 2021, 6(4): 3037-3052. doi: 10.3934/math.2021185
    [6] Ahmad A. Abubaker, Khaled Matarneh, Mohammad Faisal Khan, Suha B. Al-Shaikh, Mustafa Kamal . Study of quantum calculus for a new subclass of q-starlike bi-univalent functions connected with vertical strip domain. AIMS Mathematics, 2024, 9(5): 11789-11804. doi: 10.3934/math.2024577
    [7] Aoen, Shuhai Li, Tula, Shuwen Li, Hang Gao . New subclass of generalized close-to-convex function related with quasi-subordination. AIMS Mathematics, 2025, 10(5): 12149-12167. doi: 10.3934/math.2025551
    [8] Kholood M. Alsager, Sheza M. El-Deeb, Ala Amourah, Jongsuk Ro . Some results for the family of holomorphic functions associated with the Babalola operator and combination binomial series. AIMS Mathematics, 2024, 9(10): 29370-29385. doi: 10.3934/math.20241423
    [9] Jianhua Gong, Muhammad Ghaffar Khan, Hala Alaqad, Bilal Khan . Sharp inequalities for q-starlike functions associated with differential subordination and q-calculus. AIMS Mathematics, 2024, 9(10): 28421-28446. doi: 10.3934/math.20241379
    [10] Muhammad Sabil Ur Rehman, Qazi Zahoor Ahmad, H. M. Srivastava, Nazar Khan, Maslina Darus, Bilal Khan . Applications of higher-order q-derivatives to the subclass of q-starlike functions associated with the Janowski functions. AIMS Mathematics, 2021, 6(2): 1110-1125. doi: 10.3934/math.2021067
  • The discrete Hopfield neural network 3-satisfiability (DHNN-3SAT) model represents an innovative application of deep learning techniques to the Boolean SAT problem. Existing research indicated that the DHNN-3SAT model demonstrated significant advantages in handling 3SAT problem instances of varying scales and complexities. Compared to traditional heuristic algorithms, this model converged to local minima more rapidly and exhibited enhanced exploration capabilities within the global search space. However, the model faced several challenges and limitations. As constraints in SAT problems dynamically increased, decreased, or changed, and as problem scales expanded, the model's computational complexity and storage requirements may increase dramatically, leading to reduced performance in handling large-scale SAT problems. To address these challenges, this paper first introduced a method for designing network synaptic weights based on fundamental logical clauses. This method effectively utilized the synaptic weight information from the original SAT problem within the DHNN network, thereby significantly reducing redundant computations. Concrete examples illustrated the design process of network synaptic weights when constraints were added, removed, or updated, offering new approaches for managing the evolving constraints in SAT problems. Subsequently, the paper presented a DHNN-3SAT model optimized by genetic algorithms combined with K-modes clustering. This model employed genetic algorithm-optimized K-modes clustering to effectively cluster the initial space, significantly reducing the search space. This approach minimized the likelihood of redundant searches and reduced the risk of getting trapped in local minima, thus improving search efficiency. Experimental tests on benchmark datasets showed that the proposed model outperformed traditional DHNN-3SAT models, DHNN-3SAT models combined with genetic algorithms, and DHNN-3SAT models combined with imperialist competitive algorithms across four evaluation metrics. This study not only broadened the application of DHNN in solving 3SAT problems but also provided valuable insights and guidance for future research.



    Let A denote the class of functions f which are analytic in the open unit disk Δ={zC:|z|<1}, normalized by the conditions f(0)=f(0)1=0. So each fA has series representation of the form

    f(z)=z+n=2anzn. (1.1)

    For two analytic functions f and g, f is said to be subordinated to g (written as fg) if there exists an analytic function ω with ω(0)=0 and |ω(z)|<1 for zΔ such that f(z)=(gω)(z).

    A function fA is said to be in the class S if f is univalent in Δ. A function fS is in class C of normalized convex functions if f(Δ) is a convex domain. For 0α1, Mocanu [23] introduced the class Mα of functions fA such that f(z)f(z)z0 for all zΔ and

    ((1α)zf(z)f(z)+α(zf(z))f(z))>0(zΔ). (1.2)

    Geometrically, fMα maps the circle centred at origin onto α-convex arcs which leads to the condition (1.2). The class Mα was studied extensively by several researchers, see [1,10,11,12,24,25,26,27] and the references cited therein.

    A function fS is uniformly starlike if f maps every circular arc Γ contained in Δ with center at ζ Δ onto a starlike arc with respect to f(ζ). A function fC is uniformly convex if f maps every circular arc Γ contained in Δ with center ζ Δ onto a convex arc. We denote the classes of uniformly starlike and uniformly convex functions by UST and  UCV, respectively. For recent study on these function classes, one can refer to [7,9,13,19,20,31].

    In 1999, Kanas and Wisniowska [15] introduced the class k-UCV (k0) of k-uniformly convex functions. A function fA is said to be in the class k-UCV if it satisfies the condition

    (1+zf(z)f(z))>k|zf(z)f(z)|(zΔ). (1.3)

    In recent years, many researchers investigated interesting properties of this class and its generalizations. For more details, see [2,3,4,14,15,16,17,18,30,32,35] and references cited therein.

    In 2015, Sokół and Nunokawa [33] introduced the class MN, a function fMN if it satisfies the condition

    (1+zf(z)f(z))>|zf(z)f(z)1|(zΔ).

    In [28], it is proved that if (f)>0 in Δ, then f is univalent in Δ. In 1972, MacGregor [21] studied the class B of functions with bounded turning, a function fB if it satisfies the condition (f)>0 for zΔ. A natural generalization of the class B is B(δ1) (0δ1<1), a function fB(δ1) if it satisfies the condition

    (f(z))>δ1(zΔ;0δ1<1), (1.4)

    for details associated with the class B(δ1) (see [5,6,34]).

    Motivated essentially by the above work, we now introduce the following class k-Q(α) of analytic functions.

    Definition 1. Let k0 and 0α1. A function fA is said to be in the class k-Q(α) if it satisfies the condition

    ((zf(z))f(z))>k|(1α)f(z)+α(zf(z))f(z)1|(zΔ). (1.5)

    It is worth mentioning that, for special values of parameters, one can obtain a number of well-known function classes, some of them are listed below:

    1. k-Q(1)=k-UCV;

    2. 0-Q(α)=C.

    In what follows, we give an example for the class k-Q(α).

    Example 1. The function f(z)=z1Az(A0) is in the class k-Q(α) with

    k1b2bb(1+α)[b(1+α)+2]+4(b=|A|). (1.6)

    The main purpose of this paper is to establish several interesting relationships between k-Q(α) and the class B(δ) of functions with bounded turning.

    To prove our main results, we need the following lemmas.

    Lemma 1. ([8]) Let h be analytic in Δ with h(0)=1, β>0 and 0γ1<1. If

    h(z)+βzh(z)h(z)1+(12γ1)z1z,

    then

    h(z)1+(12δ)z1z,

    where

    δ=(2γ1β)+(2γ1β)2+8β4. (2.1)

    Lemma 2. Let h be analytic in Δ and of the form

    h(z)=1+n=mbnzn(bm0)

    with h(z)0 in Δ. If there exists a point z0(|z0|<1) such that |argh(z)|<πρ2(|z|<|z0|) and |argh(z0)|=πρ2 for some ρ>0, then z0h(z0)h(z0)=iρ, where

    :{n2(c+1c)(argh(z0)=πρ2),n2(c+1c)(argh(z0)=πρ2),

    and (h(z0))1/ρ=±ic(c>0).

    This result is a generalization of the Nunokawa's lemma [29].

    Lemma 3. ([37]) Let ε be a positive measure on [0,1]. Let ϝ be a complex-valued function defined on Δ×[0,1] such that ϝ(.,t) is analytic in Δ for each t[0,1] and ϝ(z,.) is ε-integrable on [0,1] for all zΔ. In addition, suppose that (ϝ(z,t))>0, ϝ(r,t) is real and (1/ϝ(z,t))1/ϝ(r,t) for |z|r<1 and t[0,1]. If ϝ(z)=10ϝ(z,t)dε(t), then (1/ϝ(z))1/ϝ(r).

    Lemma 4. ([22]) If 1D<C1, λ1>0 and (γ2)λ1(1C)/(1D), then the differential equation

    s(z)+zs(z)λ1s(z)+γ2=1+Cz1+Dz(zΔ)

    has a univalent solution in Δ given by

    s(z)={zλ1+γ2(1+Dz)λ1(CD)/Dλ1z0tλ1+γ21(1+Dt)λ1(CD)/Ddtγ2λ1(D0),zλ1+γ2eλ1Czλ1z0tλ1+γ21eλ1Ctdtγ2λ1(D=0).

    If r(z)=1+c1z+c2z2+ satisfies the condition

    r(z)+zr(z)λ1r(z)+γ21+Cz1+Dz(zΔ),

    then

    r(z)s(z)1+Cz1+Dz,

    and s(z) is the best dominant.

    Lemma 5. ([36,Chapter 14]) Let w, x and\ y0,1,2, be complex numbers. Then, for (y)>(x)>0, one has

    1. 2G1(w,x,y;z)=Γ(y)Γ(yx)Γ(x)10sx1(1s)yx1(1sz)wds;

    2. 2G1(w,x,y;z)= 2G1(x,w,y;z);

    3. 2G1(w,x,y;z)=(1z)w2G1(w,yx,y;zz1).

    Firstly, we derive the following result.

    Theorem 1. Let 0α<1 and k11α. If fk-Q(α), then fB(δ), where

    δ=(2μλ)+(2μλ)2+8λ4(λ=1+αkk(1α);μ=kαk1k(1α)). (3.1)

    Proof. Let f=, where is analytic in Δ with (0)=1. From inequality (1.5) which takes the form

    (1+z(z)(z))>k|(1α)(z)+α(1+z(z)(z))1|=k|1α(z)+α(z)αz(z)(z)|,

    we find that

    ((z)+1+αkk(1α)z(z)(z))>kαk1k(1α),

    which can be rewritten as

    ((z)+λz(z)(z))>μ(λ=1+αkk(1α);μ=kαk1k(1α)).

    The above relationship can be written as the following Briot-Bouquet differential subordination

    (z)+λz(z)(z)1+(12μ)z1z.

    Thus, by Lemma 1, we obtain

    1+(12δ)z1z, (3.2)

    where δ is given by (3.1). The relationship (3.2) implies that fB(δ). We thus complete the proof of Theorem 3.1.

    Theorem 2. Let 0<α1, 0<β<1, c>0, k1, nm+1(m N ), ||n2(c+1c) and

    |αβ±(1α)cβsinβπ2|1. (3.3)

    If

    f(z)=z+n=m+1anzn(am+10)

    and fk-Q(α), then fB(β0), where

    β0=min{β:β(0,1)}

    such that (3.3) holds.

    Proof. By the assumption, we have

    f(z)=(z)=1+n=mcnzn(cm0). (3.4)

    In view of (1.5) and (3.4), we get

    (1+z(z)(z))>k|(1α)(z)+α(1+z(z)(z))1|.

    If there exists a point z0Δ such that

    |arg(z)|<βπ2(|z|<|z0|;0<β<1)

    and

    |arg(z0)|=βπ2(0<β<1),

    then from Lemma 2, we know that

    z0(z0)(z0)=iβ,

    where

    ((z0))1/β=±ic(c>0)

    and

    :{n2(c+1c)(arg(z0)=βπ2),n2(c+1c)(arg(z0)=βπ2).

    For the case

    arg(z0)=βπ2,

    we get

    (1+z0(z0)(z0))=(1+iβ)=1. (3.5)

    Moreover, we find from (3.3) that

    k|(1α)(z0)+α(1+z0(z0)(z0))1|=k|(1α)((z0)1)+αz0(z0)(z0)|=k|(1α)[(±ic)β1]+iαβ|=k(1α)2(cβcosβπ21)2+[αβ±(1α)cβsinβπ2]21. (3.6)

    By virtue of (3.5) and (3.6), we have

    (1+z(z0)(z0))k|(1α)(z0)+α(1+z0(z0)(z0))1|,

    which is a contradiction to the definition of k-Q(α). Since β0=min{β:β(0,1)} such that (3.3) holds, we can deduce that fB(β0).

    By using the similar method as given above, we can prove the case

    arg(z0)=βπ2

    is true. The proof of Theorem 2 is thus completed.

    Theorem 3. If 0<β<1 and 0ν<1. If fk-Q(α), then

    (f)>[2G1(2β(1ν),1;1β+1;12)]1,

    or equivalently, k-Q(α)B(ν0), where

    ν0=[2G1(2β(1μ),1;1β+1;12)]1.

    Proof. For

    w=2β(1ν), x=1β, y=1β+1,

    we define

    ϝ(z)=(1+Dz)w10tx1(1+Dtz)wdt=Γ(x)Γ(y) 2G1(1,w,y;zz1). (3.7)

    To prove k-Q(α)B(ν0), it suffices to prove that

    inf|z|<1{(q(z))}=q(1),

    which need to show that

    (1/ϝ(z))1/ϝ(1).

    By Lemma 3 and (3.7), it follows that

    ϝ(z)=10ϝ(z,t)dε(t),

    where

    ϝ(z,t)=1z1(1t)z(0t1),

    and

    dε(t)=Γ(x)Γ(w)Γ(yw)tw1(1t)yw1dt,

    which is a positive measure on [0,1].

    It is clear that (ϝ(z,t))>0 and ϝ(r,t) is real for |z|r<1 and t[0,1]. Also

    (1ϝ(z,t))=(1(1t)z1z)1+(1t)r1+r=1ϝ(r,t)

    for |z|r<1. Therefore, by Lemma 3, we get

    (1/ϝ(z))1/ϝ(r).

    If we let r1, it follows that

    (1/ϝ(z))1/ϝ(1).

    Thus, we deduce that k-Q(α)B(ν0).

    Theorem 4. Let 0α<1 and k11α. If fk-Q(α), then

    f(z)s(z)=1g(z),

    where

    g(z)=2G1(2λ,1,1λ+1;zz1)(λ=1+αkk(1α)).

    Proof. Suppose that f=. From the proof of Theorem 1, we see that

    (z)+z(z)1λ(z)1+(12μ)z1z1+z1z(λ=1+αkk(1α);μ=kαk1k(1α)).

    If we set λ1=1λ, γ2=0, C=1 and D=1 in Lemma 4, then

    (z)s(z)=1g(z)=z1λ(1z)2λ1/λz0t(1/λ)1(1t)2/λdt.

    By putting t=uz, and using Lemma 5, we obtain

    (z)s(z)=1g(z)=11λ(1z)2λ10u(1/λ)1(1uz)2/λdu=[2G1(2λ,1,1λ+1;zz1)]1,

    which is the desired result of Theorem 4.

    The present investigation was supported by the Key Project of Education Department of Hunan Province under Grant no. 19A097 of the P. R. China. The authors would like to thank the referees for their valuable comments and suggestions, which was essential to improve the quality of this paper.

    The authors declare no conflict of interest.



    [1] M. Järvisalo, B. D. Le, O. Roussel, L. Simon, The international SAT solver competitions, Ai Mag., 33 (2012), 89–92. https://doi.org/10.1609/aimag.v33i1.2395 doi: 10.1609/aimag.v33i1.2395
    [2] S. A Cook, The complexity of theorem-proving procedures, Logic automata computat. Complex., 2023, 143152.
    [3] J. Rintanen, Planning as satisfiability: Heuristics, Artif. Intell., 193 (2012), 45–86. https://doi.org/10.1016/j.artint.2012.08.001 doi: 10.1016/j.artint.2012.08.001
    [4] V. Popov, An approach to the design of DNA smart programmable membranes, Adv. Mater. Res., 934 (2014), 173–176. https://doi.org/10.4028/www.scientific.net/AMR.934.173 doi: 10.4028/www.scientific.net/AMR.934.173
    [5] X. Zhang, J. Bussche, F. Picalausa, On the satisfiability problem for SPARQL patterns, J. Artif. Intell. Res., 56 (2016), 403–428. https://doi.org/10.1613/jair.5028 doi: 10.1613/jair.5028
    [6] A. Armando, L. Compagna, SAT-based model-checking for security protocols analysis, Int. J. Inf. Secur., 7 (2008), 3–32. https://doi.org/10.1007/s10207-007-0041-y doi: 10.1007/s10207-007-0041-y
    [7] C. Luo, S. Cai, W. Wu, K. Su, Double configuration checking in stochastic local search for satisfiability, Proc. AAAI Conf. Artif. Intell., 28 (2014), 2703–2709. https://doi.org/10.1609/aaai.v28i1.9110 doi: 10.1609/aaai.v28i1.9110
    [8] X. Wang, A novel approach of solving the CNF-SAT problem, arXiv. Prepr. arXiv., 1307 (2013), 6291. https://doi.org/10.48550/arXiv.1307.6291 doi: 10.48550/arXiv.1307.6291
    [9] D. Karaboga, B. Gorkemli, C. Ozturk, N. Karaboga, A comprehensive survey: Artificial bee colony (ABC) algorithm and applications, Artif. Intell. Rev., 42 (2014), 21–57. https://doi.org/10.1007/s10462-012-9328-0 doi: 10.1007/s10462-012-9328-0
    [10] E. A. Hirsch, A. Kojevnikov, UnitWalk: A new SAT solver that uses local search guided by unit clause elimination, Ann. Math. Artif. Intell., 43 (2005), 91–111. https://doi.org/10.1007/s10472-005-0421-9 doi: 10.1007/s10472-005-0421-9
    [11] J. J. Hopfield, Neural networks and physical systems with emergent collective computational abilities, Proc. Natl. Acad. Sci., 79 (1982), 2554–2558. https://doi.org/10.1073/pnas.79.8.2554 doi: 10.1073/pnas.79.8.2554
    [12] M. A. Mansor, S. Sathasivam, Optimal performance evaluation metrics for satisfiability logic representation in discrete Hopfield neural network, Int. J. Math. Comput. Sci., 16 (2021), 963–976.
    [13] C. C. Feng, S. Sathasivam, A novel processor for dynamic evolution of constrained SAT problems: The dynamic evolution variant of the discrete Hopfield neural network satisfiability model, J. King Saud Univ., Comput. Inf. Sci., 36 (2024), 101927. https://doi.org/10.1016/j.jksuci.2024.101927 doi: 10.1016/j.jksuci.2024.101927
    [14] S. A. Karim, M. S. M Kasihmuddin, S. Sathasivam, M. A. Mansor, S. Z. M. Jamaludin, M. R. Amin, A novel multi-objective hybrid election algorithm for higher-order random satisfiability in discrete hopfield neural network, Mathematics, 10 (2022), 1963. https://doi.org/10.3390/math10121963 doi: 10.3390/math10121963
    [15] M. A. Mansor, S. Sathasivam, Accelerating activation function for 3-satisfiability logic programming, Int.J. Intell. Syst. Appl., 8 (2016), 44. https://doi.org/10.5815/ijisa.2016.10.05. doi: 10.5815/ijisa.2016.10.05
    [16] S. Sathasivam, Upgrading logic programming in Hopfield network, Sains Malays., 39 (2010), 115–118.
    [17] M. A. Mansor, M. S. M Kasihmuddin, S. Sathasivam, Artificial immune system paradigm in the Hopfield network for 3-Satisfiability problem, Pertanika J. Sci. Technol., 25 (2017), 1173–1188.
    [18] B. Bünz, M. Lamm. Graph neural networks and boolean satisfiability, arXiv. Prepr. arXiv., 1702 (2017), 03592. https://doi.org/10.48550/arXiv.1702.03592 doi: 10.48550/arXiv.1702.03592
    [19] H. Xu, S. Koenig, T. K. S. Kumar, Towards effective deep learning for constraint satisfaction problems, Int. Conf. Princ. Pract. Constraint Program., 2018,588–597. https://doi.org/10.1007/978-3-319-98334-9_38 doi: 10.1007/978-3-319-98334-9_38
    [20] H. E. Dixon, M. L. Ginsberg, Combining satisfiability techniques from AI and OR, Knowl. Eng. Rev., 15 (2000), 31–45. https://doi.org/10.1017/S0269888900001041 doi: 10.1017/S0269888900001041
    [21] W. A. Abdullah, Logic programming on a neural network, Int. J. Intell. Syst., 7 (1992), 513–519. https://doi.org/10.1002/int.4550070604 doi: 10.1002/int.4550070604
    [22] S. Sathasivam, W. A. Abdullah, Logic mining in neural network: Reverse analysis method, Computing, 91 (2011), 119–133. https://doi.org/10.1007/s00607-010-0117-9 doi: 10.1007/s00607-010-0117-9
    [23] S. Sathasivam, N. P. Fen, M. Velavan, Reverse analysis in higher order Hopfield network for higher order horn clauses, Appl. Math. Sci., 8 (2014), 601–612. http://dx.doi.org/10.12988/ams.2014.310565 doi: 10.12988/ams.2014.310565
    [24] M. S. M. Kasihmuddin, M. A. Mansor, S. Sathasivam, Discrete Hopfield neural network in restricted maximum k-satisfiability logic programming, Sains Malays., 47 (2018), 1327–1335. http:/ldx.doi.org110.17576/jsm-2018-4706-30 doi: 10.17576/jsm-2018-4706-30
    [25] M. A. Mansor, M. S. M. Kasihmuddin, S. Sathasivam, Robust artificial immune system in the hopfield network for maximum k-satisfiability. Int. J. Inter. Mult. Artif. Intell., 4 (2017), 63–71.
    [26] M. S. M. Kasihmuddin, M. A. Mansor, S. Sathasivam, Hybrid genetic algorithm in the hopfield network for logic satisfiability problem. Pertanika J. Sci. Technol., 25 (2017), 139–152.
    [27] F. L. Azizan, S. Sathasivam M. K. M. Ali, N. Roslan, C. Feng, Hybridised network of fuzzy logic and a genetic algorithm in solving 3-Satisfiability Hopfield neural networks, Axioms, 12 (2023), 250. https://doi.org/10.3390/axioms12030250 doi: 10.3390/axioms12030250
    [28] M. A. Mansor, M. S. M. Kasihmuddin, S. Sathasivam, Grey wolf optimization algorithm with discrete hopfield neural network for 3 Satisfiability analysis, J. Phys. Conf. Ser., 1821 (2021), 012038. https://doi.org/10.1088/1742-6596/1821/1/012038 doi: 10.1088/1742-6596/1821/1/012038
    [29] M. A. Mansor, M. S. M. Kasihmuddin, S. Sathasivam, Modified lion optimization algorithm with discrete Hopfield neural network for higher order Boolean satisfiability programming, Malays. J. Math. Sci., 14 (2020), 47–61.
    [30] N. Cao, X. J. Yin, S. T. Bai, Breather wave, lump type and interaction solutions for a high dimensional evolution model, Chaos, Solitons Fract., 172 (2023), 113505. https://doi.org/10.1016/j.chaos.2023.113505 doi: 10.1016/j.chaos.2023.113505
    [31] C. C. Feng, S. Sathasivam, N. Roslan, M. Velavan, 2-SAT discrete Hopfield neural networks optimization via Crow search and fuzzy dynamical clustering approach, AIMS Math., 9 (2024), 9232–9266. https://doi.org/10.1016/10.3934/math.2024450 doi: 10.1016/10.3934/math.2024450
    [32] S. Bai, X. Yin, N. Cao, L. Xu, A high dimensional evolution model and its rogue wave solution, breather solution and mixed solutions Nonlinear Dyn., 111 (2023), 12479–12494. https://doi.org/10.1007/s11071-023-08467-x doi: 10.1007/s11071-023-08467-x
    [33] G. He, P. Tang, X. Pang, Neural network approaches to implementation of optimum multiuser detectors in code-division multiple-access channels, Int. J. Electron., 80 (1996), 425-431. https://doi.org/10.1080/002072196137264 doi: 10.1080/002072196137264
    [34] H. Yang, Z. Li, Z. Liu, A method of routing optimization using CHNN in MANET, J. Ambient Intell. Hum. Comput., 10 (2019), 1759–1768. https://doi.org/10.1007/s12652-017-0614-1 doi: 10.1007/s12652-017-0614-1
    [35] B. Bao, H. Tang, Y. Su, H. Bao, M. Chen, Q. Xu, Two-Dimensional discrete Bi-Neuron Hopfield neural network with polyhedral Hyperchaos, IEEE Trans. Circuits Syst., (2024), 1–12. https://doi.org/10.1109/TCSI.2024.3382259 doi: 10.1109/TCSI.2024.3382259
    [36] W. Ma, X. Li, T. Yu, Z. Wang, A 4D discrete Hopfield neural network-based image encryption scheme with multiple diffusion modes, Optik, 291 (2023), 171387. https://doi.org/10.1016/j.ijleo.2023.171387 doi: 10.1016/j.ijleo.2023.171387
    [37] Q. Deng, C. Wang, H. Lin, Memristive Hopfield neural network dynamics with heterogeneous activation functions and its application, Chaos Solitons Fract., 178 (2024), 114387. https://doi.org/10.1016/j.chaos.2023.114387 doi: 10.1016/j.chaos.2023.114387
    [38] Z. S. Shazli, M. B. Tahoori, Using boolean satisfiability for computing soft error rates in early design stages, Microelectr. Reliab., 50 (2010), 149–159. https://doi.org/10.1016/j.microrel.2009.08.006 doi: 10.1016/j.microrel.2009.08.006
    [39] N. Sharma, N. Gaud, K-modes clustering algorithm for categorical data, Int. J. Comput. Appl., 127 (2015), 1–6. https://doi.org/10.1016/10.5120/ijca2015906708 doi: 10.1016/10.5120/ijca2015906708
    [40] F. Cao J. Liang, D. Li, L. Bai, C. Dang, A dissimilarity measure for the K-Modes clustering algorithm, Knowl. Based Syst., 26 (2012), 120–127. https://doi.org/10.1016/j.knosys.2011.07.011 doi: 10.1016/j.knosys.2011.07.011
    [41] M. K. Ng, M. J. Li, J. Z. Huang, Z. He, On the impact of dissimilarity measure in k-modes clustering algorithm, IEEE Trans. Pattern Anal. Mach. intell., 29 (2007), 503–507. https://doi.org/10.1109/TPAMI.2007.53 doi: 10.1109/TPAMI.2007.53
    [42] S. S. Khan, A. Ahmad, Cluster center initialization algorithm for K-modes clustering, Pattern Recognit. Lett., 25 (2004), 1293–1302. https://doi.org/10.1016/j.patrec.2004.04.007 doi: 10.1016/j.patrec.2004.04.007
    [43] R. J. Kuo, Y. R. Zheng, T. P. Q. Nguyen, Metaheuristic-based possibilistic fuzzy k-modes algorithms for categorical data clustering, Inf. Sci., 557 (2021), 1–15. https://doi.org/10.1016/j.ins.2020.12.051 doi: 10.1016/j.ins.2020.12.051
    [44] G. Gan, J. Wu, Z. Yang, A genetic fuzzy k-Modes algorithm for clustering categorical data, Expert Syst. Appl., 36 (2009), 1615–1620. https://doi.org/10.1007/11527503_23 doi: 10.1007/11527503_23
    [45] F. S. Gharehchopogh, S. Haggi, An Optimization K-modes clustering algorithm with elephant herding optimization algorithm for crime clustering, J. Adv. Comput. Eng. Technol., 6 (2020), 79–90.
    [46] J. Gu, Local search for satisfiability (SAT) problem, IEEE Trans. Syst., 23 (1993), 1108–1129. https://doi.org/10.1109/21.247892 doi: 10.1109/21.247892
    [47] N. E. Zamri, M. A, Mansor, M. S. M. Kasihmuddin, A. Always, S. A. Alzaeemi, Amazon employees resources access data extraction via clonal selection algorithm and logic mining approach, Entropy, 22 (2020), 596. https://doi.org/10.3390/e22060596 doi: 10.3390/e22060596
  • This article has been cited by:

    1. A Certain Class of Function Analytic and Subordinate to the Modified Sigmoid Function, 2025, 2581-8147, 639, 10.34198/ejms.15425.639647
    2. Tamer M. Seoudy, Amnah E. Shammaky, Certain subfamily of multivalently Bazilevič and non-Bazilevič functions involving the bounded boundary rotation, 2025, 10, 2473-6988, 12745, 10.3934/math.2025574
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1011) PDF downloads(43) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog