
With the widespread application of smart devices, the security of internet of things (IoT) systems faces entirely new challenges. The IoT data stream operates in a non-stationary, dynamic environment, making it prone to concept drift. This paper focused on addressing the issue of concept drift in data streams, with a key emphasis on introducing an innovative drift detection method-ensemble multiple non-parametric concept localization detectors, abbreviated as EMNCD. EMNCD employs an ensemble of non-parametric statistical methods, including the Kolmogorov-Smirnov, Wilcoxon rank sum and Mann-Kendall tests. By comparing sample distributions within a sliding window, EMNCD accurately detects concept drift, achieving precise localization of drift points, and enhancing overall detection reliability. Experimental results demonstrated the superior performance of EMNCD compared to classical methods on artificial datasets. Simultaneously, to enhance the robustness of data stream processing, we presented an online anomaly detection method based on the isolation forest (iForest). Additionally, we proposedwhale optimization algorithm (WOA)-extreme gradient boosting (XGBoost), a drift adaptation model employing XGBoost as a base classifier. This model dynamically updates using drift points detected by EMNCD and fine-tunes parameters through the WOA. Real-world applications on the edge-industrial IoTset (IIoTset) intrusion dataset explore the impact of concept drift on intrusion detection, where IIoT is a subclass of IoT. In summary, this paper focused on EMNCD, introducing innovative approaches for drift detection, anomaly detection, and drift adaptation. The research provided practical and viable solutions to address concept drift in data streams, enhancing security in IoT systems.
Citation: Renjie Chu, Peiyuan Jin, Hanli Qiao, Quanxi Feng. Intrusion detection in the IoT data streams using concept drift localization[J]. AIMS Mathematics, 2024, 9(1): 1535-1561. doi: 10.3934/math.2024076
[1] | Ali Al Khabyah, Haseeb Ahmad, Ali Ahmad, Ali N. A. Koam . A uniform interval-valued intuitionistic fuzzy environment: topological descriptors and their application in neural networks. AIMS Mathematics, 2024, 9(10): 28792-28812. doi: 10.3934/math.20241397 |
[2] | Tareq Saeed . Intuitionistic fuzzy variational inequalities and their applications. AIMS Mathematics, 2024, 9(12): 34289-34310. doi: 10.3934/math.20241634 |
[3] | Rashid Ali, Faisar Mehmood, Aqib Saghir, Hassen Aydi, Saber Mansour, Wajdi Kallel . Solution of integral equations for multivalued maps in fuzzy b-metric spaces using Geraghty type contractions. AIMS Mathematics, 2023, 8(7): 16633-16654. doi: 10.3934/math.2023851 |
[4] | Mohammed Shehu Shagari, Saima Rashid, Fahd Jarad, Mohamed S. Mohamed . Interpolative contractions and intuitionistic fuzzy set-valued maps with applications. AIMS Mathematics, 2022, 7(6): 10744-10758. doi: 10.3934/math.2022600 |
[5] | Changlin Xu, Yaqing Wen . New measure of circular intuitionistic fuzzy sets and its application in decision making. AIMS Mathematics, 2023, 8(10): 24053-24074. doi: 10.3934/math.20231226 |
[6] | Doaa Al-Sharoa . (α1, 2, β1, 2)-complex intuitionistic fuzzy subgroups and its algebraic structure. AIMS Mathematics, 2023, 8(4): 8082-8116. doi: 10.3934/math.2023409 |
[7] | Zhihua Wang . Stability of a mixed type additive-quadratic functional equation with a parameter in matrix intuitionistic fuzzy normed spaces. AIMS Mathematics, 2023, 8(11): 25422-25442. doi: 10.3934/math.20231297 |
[8] | Rana Muhammad Zulqarnain, Xiao Long Xin, Muhammad Saeed . Extension of TOPSIS method under intuitionistic fuzzy hypersoft environment based on correlation coefficient and aggregation operators to solve decision making problem. AIMS Mathematics, 2021, 6(3): 2732-2755. doi: 10.3934/math.2021167 |
[9] | Samina Batul, Faisar Mehmood, Azhar Hussain, Dur-e-Shehwar Sagheer, Hassen Aydi, Aiman Mukheimer . Multivalued contraction maps on fuzzy b-metric spaces and an application. AIMS Mathematics, 2022, 7(4): 5925-5942. doi: 10.3934/math.2022330 |
[10] | Afrah Ahmad Noman Abdou . Chatterjea type theorems for complex valued extended b-metric spaces with applications. AIMS Mathematics, 2023, 8(8): 19142-19160. doi: 10.3934/math.2023977 |
With the widespread application of smart devices, the security of internet of things (IoT) systems faces entirely new challenges. The IoT data stream operates in a non-stationary, dynamic environment, making it prone to concept drift. This paper focused on addressing the issue of concept drift in data streams, with a key emphasis on introducing an innovative drift detection method-ensemble multiple non-parametric concept localization detectors, abbreviated as EMNCD. EMNCD employs an ensemble of non-parametric statistical methods, including the Kolmogorov-Smirnov, Wilcoxon rank sum and Mann-Kendall tests. By comparing sample distributions within a sliding window, EMNCD accurately detects concept drift, achieving precise localization of drift points, and enhancing overall detection reliability. Experimental results demonstrated the superior performance of EMNCD compared to classical methods on artificial datasets. Simultaneously, to enhance the robustness of data stream processing, we presented an online anomaly detection method based on the isolation forest (iForest). Additionally, we proposedwhale optimization algorithm (WOA)-extreme gradient boosting (XGBoost), a drift adaptation model employing XGBoost as a base classifier. This model dynamically updates using drift points detected by EMNCD and fine-tunes parameters through the WOA. Real-world applications on the edge-industrial IoTset (IIoTset) intrusion dataset explore the impact of concept drift on intrusion detection, where IIoT is a subclass of IoT. In summary, this paper focused on EMNCD, introducing innovative approaches for drift detection, anomaly detection, and drift adaptation. The research provided practical and viable solutions to address concept drift in data streams, enhancing security in IoT systems.
A problem that occurs frequently in a variety of mathematical contexts, is to find the common invariant subspaces of a single matrix or set of matrices. In the case of a single endomorphism or matrix, it is relatively easy to find all the invariant subspaces by using the Jordan normal form. Also, some theoretical results are given only for the invariant subspaces of two matrices. However, when there are more than two matrices, the problem becomes much harder, and unexpected invariant subspaces may occur. No systematic method is known. In a recent article [1], we have provided a new algorithms to determine common invariant subspaces of a single matrix or of a set of matrices systematically.
In the present article we consider a more general version of this problem, that is, providing two algorithms for simultaneous block triangularization and block diagonalization of sets of matrices. One of the main steps in the first two proposed algorithms, consists of finding the common invariant subspaces of matrices using the new method proposed in the recent article [1]. It is worth mentioning that an efficient algorithm to explicitly compute a transfer matrix which realizes the simultaneous block diagonalization of unitary matrices whose decomposition in irreducible blocks (common invariant subspaces) is known from elsewhere is given in [2]. An application of simultaneous block-diagonalization of normal matrices in quantum theory is presented in [3].
In this article we shall be concerned with finite dimensions only. Of course the fact that a single complex matrix can always be put into triangular form follows readily from the Jordan normal form theorem [4]. For a set of matrices, Jacobson in [5] introduced the notion of a composition series for a collection of matrices. The idea of a composition series for a group is quite familiar. The Jordan-Hölder Theorem [4] states that any two composition series of the same group have the same length and the same composition factors (up to permutation). Jacobson in [5] characterized the simultaneous block triangularization of a set of matrices by the existence of a chain {0}=V0⊂V1⊂...⊂Vt=Cn of invariant subspaces with dimension dim(Vi/Vi−1)=ni. Therefore, in the context of a collection of matrices Ω={Ai}Ni=1, the idea is to locate a common invariant subspace V of minimal dimension d of a set of matrices Ω. Assume V is generated by the (linearly independent) set B1={u1,u2,...,ud}, and let B={u1,u2,...,ud,ud+1,ud+2,...,un} be a basis of Cn containing B1. Upon setting S=(u1,u2,...,ud,ud+1,ud+2,...,un), S−1AiS has the block triangular form
S−1AiS=(Bi1,1Bi1,20Bi2,2), |
for i=1,...,n. Thereafter, one may define a quotient of the ambient vector space, and each of the matrices in the given collection will pass to this quotient. As such, one defines
Ti=Bi2,2=(0(n−d)×dIn−d)S−1AiS(0d×(n−d)In−d). |
Then one may begin again the process of looking for a common invariant subspace of minimal dimension of a set of matrices {Ti}Ni=1 and iterate the procedure. Since all spaces and matrices are of finite dimension, the procedure must terminate at some point. Again, any two such composition series will be isomorphic. When the various quotients and submatrices are lifted back to the original vector space, one obtains precisely the block-triangular form for the original set of matrices. It is important to find a composition series in the construction in order to make the set of matrices as "block-triangular as possible."
Dubi [6] gave an algorithmic approach to simultaneous triangularization of a set of matrices based on the idea of Jacobson in [5]. In the case of simultaneous triangularization, it can be understood as the existence of a chain {0}=V0⊂V1⊂...⊂Vt=Cn of invariant subspaces with dimension dim(Vi)=i. We generalize his study to cover simultaneous block triangularization of a set of matrices. The generalized algorithm depends on the novel algorithm for constructing invariant subspaces of a set of matrices given in the recent article [1].
Specht [7] (see also [8]) proved that if the associative algebra L generated by a set of matrices Ω over C satisfies L=L∗, then Ω admits simultaneous block triangularization if and only if it admits simultaneous block diagonalization, in both cases via a unitary matrix. Following a result of Specht, we prove that a set of matrices Ω admits simultaneous block diagonalization if and only if the set Γ=Ω∪Ω∗ admits simultaneous block triangularization. Finally, an algorithmic approach to simultaneous block diagonalization of a set of matrices based on this fact is proposed.
The latter part of this paper presents an alternate approach for simultaneous block diagonalization of a set of n×n matrices {As}Ns=1 by an invertible matrix that does not require finding the common invariant subspaces. Maehara et al. [9] introduced an algorithm for simultaneous block diagonalization of a set of matrices by a unitary matrix based on the existence of a Hermitian commuting matrix. Here, we extend their algorithm to simultaneous block diagonalization of a set of matrices by an invertible matrix based on the existence of a commuting matrix which is not necessarily Hermitian. For example, consider the set of matrices Ω={Ai}2i=1 where
A1=(100220111),A2=(000210010). | (1.1) |
The only Hermitian matrix commuting with the set Ω is the identity matrix. Therefore, we cannot apply the proposed algorithm given in [9]. However, one can verify that the following non Hermitian matrix C commutes with all the matrices {Ai}2i=1
C=(000210010). | (1.2) |
The matrix C has distinct eigenvalues λ1=0,λ2=1 with algebraic multiplicities n1=2,n2=1, respectively. Moreover, the matrix C is not diagonalizable. Therefore, we cannot construct the eigenvalue decomposition for the matrix C. However, one can decompose the matrix C by its generalized eigen vectors as follows:
S−1CS=(010000001)=(0100)⊕(1), | (1.3) |
where
S=(0−120011101). | (1.4) |
Initially, it is noted that the matrices {Ai}2i=1 can be decomposed into two diagonal blocks by the constructed invertible matrix S where
S−1A1S=(11201)⊕(2),S−1A2S=(0100)⊕(1). | (1.5) |
Then, a new algorithm is developed for simultaneous block diagonalization by an invertible matrix based on the generalized eigenvectors of a commuting matrix. Moreover, a new characterization is presented by proving that the existence of a commuting matrix that possesses at least two distinct eigenvalues is the necessary and sufficient condition to guarantee the simultaneous block diagonalization by an invertible matrix.
An outline of the paper is as follows. In Section 2 we review several definitions pertaining to block-triangular and block-diagonal matrices and state several elementary consequences that follow from them. In Section 3, following a result of Specht [7] (see also [8]), we provide conditions for putting a set of matrices into block-diagonal form simultaneously. Furthermore, we apply the theoretical results to provide two algorithms that enable a collection of matrices to be put into block-triangular form or block-diagonal form simultaneously by a unitary matrix based on the existence of invariant subspaces. In Section 4, a new characterization is presented by proving that the existence of a commuting matrix that possesses at least two distinct eigenvalues is the necessary and sufficient condition to guarantee the simultaneous block diagonalization by an invertible matrix. Furthermore, we apply the theoretical results to provide an algorithm that enables a collection of matrices to be put into block-diagonal form simultaneously by an invertible matrix based on the existence of a commuting matrix. Sections 3 and 4 also provide concrete examples using the symbolic manipulation system Maple.
Let Ω be a set of n×n matrices over an algebraically closed field F, and let L denote the algebra generated by Ω over F. Similarly, let Ω∗ be the set of the conjugate transpose of each matrix in Ω and L∗ denote the algebra generated by Ω∗ over F.
Definition 2.1. An n×n matrix A is given the notation BT(n1,...,nt) provided A is block upper triangular with t square blocks on the diagonal, of sizes n1,...,nt, where t≥2 and n1+...+nt=n. That is, a block upper triangular matrix A has the form
A=(A1,1A1,2⋯A1,t0A2,2⋯A2,t⋮⋮⋱⋮00⋯At,t) | (2.1) |
where Ai,j is a square matrix for all i=1,...,t and j=i,...,t.
Definition 2.2. A set of n×n matrices Ω is BT(n1,...,nt) if all of the matrices in Ω are BT(n1,...,nt).
Remark 2.3. A set of n×n matrices Ω admits a simultaneous triangularization if it is BT(n1,...,nt) with ni=1 for i=1,...,t.
Remark 2.4. A set of n×n matrices Ω is BT(n1,...,nt) if and only if the algebra L generated by Ω is BT(n1,...,nt).
Proposition 2.5. [7] (see also [8]) Let Ω be a nonempty set of complex n×n matrices. Then, there is a nonsingular matrix S such that SΩS−1 is BT(n1,...,nt) if and only if there is a unitary matrix U such that UΩU∗ is BT(n1,...,nt).
Theorem 2.6. [5,Chapter Ⅳ] Let Ω be a nonempty set of complex n×n matrices. Then, there is a unitary matrix U such that UΩU∗ is BT(n1,...,nt) if and only if the set Ω has a chain {0}=V0⊂V1⊂...⊂Vt=Cn of invariant subspaces with dimension dim(Vi/Vi−1)=ni.
Definition 2.7. An n×n matrix A is given the notation BD(n1,...,nt) provided A is block diagonal with t square blocks on the diagonal, of sizes n1,...,nt, where t≥2, n1+...+nt=n, and the blocks off the diagonal are the zero matrices. That is, a block diagonal matrix A has the form
A=(A10⋯00A2⋯0⋮⋮⋱⋮00⋯At) | (2.2) |
where Ak is a square matrix for all k=1,...,t. In other words, matrix A is the direct sum of A1,...,At. It can also be indicated as A1⊕A2⊕...⊕At.
Definition 2.8. A set of n×n matrices Ω is BD(n1,...,nt) if all of the matrices in Ω are BD(n1,...,nt).
Remark 2.9. A set of n×n matrices Ω admits a simultaneous diagonalization if it is BD(n1,...,nt) with ni=1 for i=1,...,t.
Remark 2.10. A set of n×n matrices Ω is BD(n1,...,nt) if and only if the algebra L generated by Ω is BD(n1,...,nt).
Proposition 2.11. [7] (see also [8]) Let Ω be a nonempty set of complex n×n matrices and let L be the algebra generated by Ω over C. Suppose L=L∗. Then, there is a nonsingular matrix S such that SLS−1 is BT(n1,...,nt) if and only if there is a unitary matrix U such that ULU∗ is BD(n1,...,nt).
Dubi [6] gave an algorithmic approach to simultaneous triangularization of a set of n×n matrices. In this section, we will generalize his study to cover simultaneous block triangularization and simultaneous block diagonalization of a set of n×n matrices. The generalized algorithms depend on the novel algorithm for constructing invariant subspaces of a set of matrices given in the recent article [1] and Theorem 3.3.
Lemma 3.1. Let Ω be a nonempty set of complex n×n matrices, Ω∗ be the set of the conjugate transpose of each matrix in Ω and L be the algebra generated by Γ=Ω∪Ω∗. Then, L=L∗.
Proof. Let A be a matrix in L. Then, A=P(B1,...,Bm) for some multivariate noncommutative polynomial P(x1,...,xm) and matrices {Bi}mi=1∈Γ. Therefore, A∗=P∗(B1,...,Bm)=Q(B∗1,...,B∗m) for some multivariate noncommutative polynomial Q(x1,...,xm) where the matrices {B∗i}mi=1∈Γ∗=Γ. Hence, the matrix A∗∈L
Lemma 3.2. Let Ω be a nonempty set of complex n×n matrices and Ω∗ be the set of the conjugate transpose of each matrix in Ω, and Γ=Ω∪Ω∗. Then, there is a unitary matrix U such that UΓU∗ is BD(n1,...,nt) if and only if there is a unitary matrix U such that UΩU∗ is BD(n1,...,nt).
Proof. Assume that there exists a unitary matrix U such that UΩU∗ is BD(n1,...,nt). Then, (UΩU∗)∗=UΩ∗U∗ is BD(n1,...,nt). Hence, UΓU∗ is BD(n1,...,nt).
Theorem 3.3. Let Ω be a nonempty set of complex n×n matrices and Ω∗ be the set of the conjugate transpose of each matrix in Ω, and Γ=Ω∪Ω∗. Then, there is a unitary matrix U such that UΩU∗ is BD(n1,...,nt) if and only if there is a unitary matrix U such that UΓU∗ is BT(n1,...,nt).
Proof. Let L be the algebra generated by Γ. Then, L=L∗ using Lemma 3.1. Now, by applying Proposition 2.11 and Lemma 3.2, the following statements are equivalent :
There is a unitary matrix U such that UΓU∗ is BT(n1,...,nt).
⟺ There is a unitary matrix U such that ULU∗ is BT(n1,...,nt).
⟺ There is a unitary matrix U such that ULU∗ is BD(n1,...,nt).
⟺ There is a unitary matrix U such that UΓU∗ is BD(n1,...,nt).
⟺ There is a unitary matrix U such that UΩU∗ is BD(n1,...,nt).
(1) Input: the set Ω={Ai}Ni=1.
(2) Set k=0,B=ϕ,s=n,Ti=Ai,S2=I.
(3) Search for a d-dimensional invariant subspace V=⟨v1,v2,...,vd⟩ of a set of matrices {Ti}Ni=1 starting from d=1 up to d=s−1. If one does not exist and k=0, abort and print "no simultaneous block triangularization". Else, if one does not exist and k≠0, go to step (8). Else, go to next step.
(4) Set Vk+1=(S2v1S2v2...S2vd),B=B∪{S2v1,S2v2,...,S2vd},S1=(V1V2...Vk+1).
(5) Find a basis {u1,u2,...,ul} for the orthogonal complement of B.
(6) Set S2=(u1u2...ul),S=(S1S2), and
Ti=(0(s−d)×dIs−d)S−1AiS(0d×(s−d)Is−d).
(7) Set k=k+1,s=s−d, and return to step (3).
(8) Compute the QR decomposition of the invertible matrix S, by means of the Gram–Schmidt process, to convert it to a unitary matrix Q.
(9) Output: a unitary matrix U as the conjugate transpose of the resulting matrix Q.
Remark 3.4. If one uses any non-orthogonal complement in step 5 of Algorithm A, then the matrix S is invertible such that S−1ΩS is BT(n1,...,nt). However, in such a case, one cannot guarantee that UΩU∗ is BT(n1,...,nt).
Example 3.5. The set of matrices Ω={Ai}2i=1 admits simultaneous block triangularization where
A1=(321011050000014012131113020025010006),A2=(44124−484036000−1012320444168524404−102880400040). | (3.1) |
Applying Algorithm A to the set Ω can be summarized as follows:
● Input: Ω.
● Initiation step:
We have k=0,B=ϕ,s=6,T1=A1,T2=A2,S2=I.
● In the first iteration:
We found two-dimensional invariant subspace V=⟨e1,e4⟩ of a set of matrices {Ti}2i=1. Therefore, B={e1,e4},S1=(e1,e4),S2=(e2,e3,e5,e6),
T1=(5000141220251006),T2=(3600−11232444−128840040), | (3.2) |
k=1, and s=4.
● In the second iteration: We found two-dimensional invariant subspace V=⟨e2,e3⟩ of a set of matrices {Ti}2i=1. Therefore, B={e1,e4,e3,e5},S1=(e1,e4,e3,e5),S2=(e2,e6),
T1=(5016),T2=(36−1440), | (3.3) |
k=2, and s=2.
● In the third iteration: There is no one-dimensional invariant subspace of a set of matrices {Ti}2i=1. Therefore, S=(e1e4e3e5e2e6), and the corresponding unitary matrix is
U=(100000000100001000000010010000000001) |
such that the set UΩU∗={UAiU∗}2i=1 is BT(2,2,2) where
UA1U∗=(301121111133004112000225000050000016),UA2U∗=(44−448124452841640032412400−12848000036−10000440). | (3.4) |
(1) Input: the set Ω={Ai}Ni=1.
(2) Construct the set Γ=Ω∪Ω∗.
(3) Find a unitary matrix U such that UΓU∗ is BT(n1,...,nt) using Algorithm A.
(4) Output: a unitary matrix U.
Remark 3.6. Algorithm B provides the finest block-diagonalization. Moreover, the number of the blocks equals the number the of the invariant subspaces, and the size of each block is ni×ni, where ni is the dimension of the invariant subspace.
Example 3.7. The set of matrices Ω={Ai}2i=1 admits simultaneous block diagonalization where
A1=(3000000020000000200000001000000010000000100000003),A2=(0000000000000001000000000000000000000010001000000). | (3.5) |
Applying Algorithm B to the set Ω can be summarized as follows:
● Input: Γ=Ω∪Ω∗.
● Initiation step:
We have k=0,B=ϕ,s=7,T1=A1,T2=A2,T3=AT2,S2=I.
● In the first iteration:
We found one-dimensional invariant subspace V=⟨e5⟩ of a set of matrices {Ti}3i=1. Therefore, B={e5},S1=(e5),S2=(e1,e2,e3,e4,e6,e7),
T1=(300000020000002000000100000010000003),T2=(000000000000010000000000000100100000),T3=TT2, | (3.6) |
k=1, and s=6.
● In the second iteration: We found two-dimensional invariant subspace V=⟨e4,e5⟩ of a set of matrices {Ti}3i=1. Therefore, B={e5,e4,e6},S1=(e5e4e6),S2=(e1,e2,e3,e7),
T1=(3000020000200003),T2=(0000000001001000),T3=TT2, | (3.7) |
k=2, and s=4.
● In the third iteration: We found two-dimensional invariant subspace V=⟨e2,e3⟩ of a set of matrices {Ti}3i=1. Therefore, B={e5,e4,e6,e2,e3},S1=(e5e4e6e2e3),S2=(e1,e7),
T1=(3003),T2=(0010),T3=(0100), | (3.8) |
k=3, and s=2.
● In the fourth iteration: There is no one-dimensional invariant subspace of a set of matrices {Ti}3i=1. Therefore, S=(e5e4e6e2e3e1e7), and the corresponding unitary matrix is
U=(0000100000100000000100100000001000010000000000001) |
such that the set UΩU∗={UAiU∗}2i=1 is BD(1,2,2,2) where
UA1U∗=(1)⊕(1001)⊕(2002)⊕(3003),UA2U∗=(0)⊕(0010)⊕(0010)⊕(0010). | (3.9) |
Example 3.8. The set of matrices Ω={Ai}2i=1 admits simultaneous block diagonalization where
A1=(3000000020000000200000001000000010000000100000003),A2=(0000000000100001000000000000000010000001001000000). | (3.10) |
Similarly, applying Algorithm B to the set Ω provides the matrix S=(e6e5e7e1e3e2e4). Therefore, the corresponding unitary matrix is
U=(0000010000010000000011000000001000001000000001000) |
such that the set UΩU∗={UAiU∗}2i=1 is BD(2,2,3) where
UA1U∗=(1001)⊕(3003)⊕(200020001),UA2U∗=(0101)⊕(0100)⊕(010001000). | (3.11) |
Example 3.9. The set of matrices Ω={Ai}3i=1 admits simultaneous block diagonalization where
A1=(000000000020000000001000000000−20000000000000000000−1000000000−100000000010000000000),A2=(000100000−100010000000001000000000000000−100000000000000000000000000000−100000000000),A3=(0−100000000000000000000000001000−100000100000000010000000000000−10000000000000000000). | (3.12) |
Similarly, applying Algorithm B to the set Ω provides the matrix S=(e1+e5e9e3e6e8−e7e1−e5,e2e4). Therefore, the corresponding unitary matrix is
U=(12√200012√20000000000001001000000000001000000000010000000−10012√2000−12√20000010000000000100000) |
such that the set UΩU∗={UAiU∗}3i=1 is BD(1,1,2,2,3) where
UA1U∗=(0)⊕(0)⊕(100−1)⊕(100−1)⊕(00002000−2),UA2U∗=(0)⊕(0)⊕(0100)⊕(0100)⊕(00√2−√200000),UA3U∗=(0)⊕(0)⊕(0010)⊕(0010)⊕(0−√20000√200). | (3.13) |
This section focuses on an alternate approach for simultaneous block diagonalization of a set of n×n matrices {As}Ns=1 by an invertible matrix that does not require finding the common invariant subspaces as Algorithm B given in the previous section. Maehara et al. [9] introduced an algorithm for simultaneous block diagonalization of a set of matrices by a unitary matrix based on the eigenvalue decomposition of a Hermitian commuting matrix. Here, we extend their algorithm to be applicable for a non-Hermitian commuting matrix by considering its generalized eigen vectors. Moreover, a new characterization is presented by proving that the existence of a commuting matrix that possesses at least two distinct eigenvalues is the necessary and sufficient condition to guarantee the simultaneous block diagonalization by an invertible matrix.
Proposition 4.1. Let V be a vector space, and let T:V→V be a linear operator. Let λ1,...,λk be distinct eigenvalues of T. Then, each generalized eigenspace Gλi(T) is T-invariant, and we have the direct sum decomposition
V=Gλ1(T)⊕Gλ2(T)⊕...⊕Gλk(T). |
Lemma 4.2. Let V be a vector space, and let T:V→V, L:V→V be linear commuting operators. Let λ1,...,λk be distinct eigenvalues of T. Then, each generalized eigenspace Gλi(T) is L-invariant.
Proof. Let V be a vector space and λ1,...,λk be distinct eigenvalues of T with the minimal polynomial μ(x)=(x−λ1)n1(x−λ2)n2...(x−λk)nk. Then, we have the direct sum decomposition V=Gλ1(T)⊕Gλ2(T)⊕...⊕Gλk(T).
For each i=1,..,k, let x∈Gλi(T), and then (T−λiI)nix=0. Then, (T−λiI)niLx=L(T−λiI)nix=0. Hence, Lx∈Gλi(T).
Theorem 4.3. Let {As}Ns=1 be a set of n×n matrices. Then, the set {As}Ns=1 admits simultaneous block diagonalization by an invertible matrix S if and only if the set {As}Ns=1 commutes with a matrix C that possesses two distinct eigenvalues.
Proof. ⇒ Assume that the set {As}Ns=1 admits simultaneous block diagonalization by the an invertible matrix S such that
S−1AsS=Bs,1⊕Bs,2⊕...⊕Bs,k, |
where the number of blocks k≥2, and the matrices Bs,1,Bs,2,...,Bs,k have sizes n1×n1,n2×n2,...,nk×nk, respectively, for all s=1,..,N.
Now, define the matrix C as
C=S(λ1In1×n1⊕λ2In2×n2⊕...⊕λkInk×nk)S−1, |
where λ1,λ2,...,λk are any distinct numbers.
Clearly, the matrix C commutes with the set {As}Ns=1. Moreover, it has the distinct eigenvalues λ1,λ2,...,λk.
⇐ Assume that the set {As}Ns=1 commutes with a matrix C that posseses distinct eigenvalues λ1,λ2,...,λk.
Using Proposition 4.1, one can use the generalized eigenspace Gλi(C) of the matrix C associated to these distinct eigenvalues to decompose the matrix C as a direct sum of k matrices. This can be achieved by restricting the matrix C on the invariant subspaces Gλi(C) as follows:
S−1CS=[C]Gλ1(C)⊕[C]Gλ2(C)⊕...⊕[C]Gλk(C) |
where
S=(Gλ1(C),Gλ2(C),...,Gλk(C)). |
Using Lemma 4.2, one can restrict each matrix As on the invariant subspaces Gλi(C) to decompose the matrix As as a direct sum of k matrices as follows:
S−1AsS=[As]Gλ1(C)⊕[As]Gλ2(C)⊕...⊕[As]Gλk(C). |
Remark 4.4. For a given set of n×n matrices {As}Ns=1, if the set {As}Ns=1 commutes only with the matrices having only one eigenvalue, then it does not admit a simultaneous block diagonalization by an invertible matrix.
Algorithm C:
(1) Input: the set Ω={As}Ns=1.
(2) Construct the the following matrix:
X=(I⊗A1−AT1⊗II⊗A2−AT2⊗I...I⊗AN−ATN⊗I). |
(3) Compute the null space of the matrix X and reshape the obtained vectors as n×n matrices. These matrices commute with all the matrices {As}Ns=1.
(4) Choose a matrix C from the obtained matrices that possesses two distinct eigenvalues.
(5) Find the distinct eigenvalues λ1,...,λk of the matrix C and the corresponding algebraic multiplicity n1,n2,...,nk.
(6) Find each generalized eigenspace Gλi(C) of the matrix C associated to the eigenvalue λi by computing the null space of (C−λiI)ni.
(7) Construct the invertible matrix S as
S=(Gλ1(C),Gλ2(C),...,Gλk(C)). |
(8) Verify that
S−1AsS=Bs,1⊕Bs,2⊕...⊕Bs,k, |
where the matrices Bs,1,Bs,2,...,Bs,k have sizes n1×n1,n2×n2,...,nk×nk, respectively, for all s=1,..,N.
(9) Output: an invertible matrix S.
Remark 4.5. Algorithm C provides the finest block-diagonalization if one chooses a matrix C with maximum number of distinct eigenvalues. Moreover, the number of the blocks equals the number the of the distinct eigenvalues, and the size of each block is ni×ni, where ni is the algebraic multiplicity of the eigenvalue λi.
Example 4.6. Consider the set of matrices Ω={Ai}6i=1 where
A1=(0000000001000000100−1000000−1000000000),A2=(000−10000000000000110000000000000−1000),A3=(0000−1000000−1000000000000100000010000),A4=(010000−1000000000000000000000010000−10),A5=(001000000000−10000000000−1000000000100),A6=(0000000010000−10000000010000−100000000). | (4.1) |
The set Ω admits simultaneous block diagonalization by an invertible matrix. An invertible matrix can be obtained by applying algorithm C to the set Ω as summarized below:
● A matrix C that commutes with all the matrices {Ai}6i=1 can be obtained as
C=(0000010000−100001000010000−10000100000). | (4.2) |
.
● The distinct eigenvalues of the matrix C are λ1=−1,λ2=1 with algebraic multiplicities n1=3,n2=3, respectively..
● The generalized eigenspaces of the matrix C associated to the distinct eigenvalues are
Gλ1(C)=N(C−λ1I)3=⟨e6−e1,e2+e5,e4−e3⟩,Gλ2(C)=N(C−λ2I)3=⟨e1+e6,e5−e2,e3+e4⟩. | (4.3) |
● The invertible matrix S=(Gλ1(C),Gλ2(C)) is
S=(−1001000100−1000−1001001001010010100100). | (4.4) |
● The set S−1ΩS={S−1AiS}6i=1 contains block diagonal matrices where
S−1A1S=(0000010−10)⊕(00000−1010),S−1A2S=(001000−100)⊕(00−1000100),S−1A3S=(010−100000)⊕(0−10100000),S−1A4S=(0−10100000)⊕(0−10100000),S−1A5S=(001000−100)⊕(001000−100),S−1A6S=(00000−1010)⊕(00000−1010). | (4.5) |
It is well known that a set of non-defective matrices can be simultaneously diagonalized if and only if the matrices commute. In the case of non-commuting matrices, the best that can be achieved is simultaneous block diagonalization. Both Algorithm B and the Maehara et al. [9] algorithm are applicable for simultaneous block diagonalization of a set of matrices by a unitary matrix. Algorithm C can be applied for block diagonalization by an invertible matrix when finding a unitary matrix is not possible. In case block diagonalization of a set of matrices is not possible by a unitary or an invertible matrix, then one may utilize block triangularization by Algorithm A. Algorithms A and B are based on the existence of invariant subspaces; however, Algorithm C is based on the existence of a commuting matrix which is not necessarily Hermitian, unlike the Maehara et al. algorithm.
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
Ahmad Y. Al-Dweik and M. T. Mustafa would like to thank Qatar University for its support and excellent research facilities. R. Ghanam and G. Thompson are grateful to VCU Qatar and Qatar Foundation for their support.
The authors declare that they have no conflicts of interest.
[1] |
M. A. Hoque, C. Davidson, Design and implementation of an IoT-based smart home security system, Int. J. Networked Distrib. Comput., 7 (2019), 85–92. https://doi.org/10.2991/ijndc.k.190326.004 doi: 10.2991/ijndc.k.190326.004
![]() |
[2] |
Z. Chen, C. Sivaparthipan, B. Muthu, IoT based smart and intelligent smart city energy optimization, Sustainable Energy Technol. Assess., 49 (2022), 101724. https://doi.org/10.1016/j.seta.2021.101724 doi: 10.1016/j.seta.2021.101724
![]() |
[3] |
D. Dobrilović, M. M, D. Malić, Learning platform for smart city application development, Interdiscip. Descr. Complex Syst., 17 (2019), 430–437. https://doi.org/10.7906/indecs.17.3.1 doi: 10.7906/indecs.17.3.1
![]() |
[4] |
R. W. Liu, Y. Guo, Y. Lu, K. T. Chui, B. B. Gupta, Deep network-enabled haze visibility enhancement for visual IoT-driven intelligent transportation systems, IEEE Trans. Ind. Inf., 19 (2022), 1581–1591. https://doi.org/10.1109/TII.2022.3170594 doi: 10.1109/TII.2022.3170594
![]() |
[5] |
X. Zhan, W. Wu, L. Shen, W. Liao, Z. Zhao, J. Xia, Industrial internet of things and unsupervised deep learning enabled real-time occupational safety monitoring in cold storage warehouse, Safety Sci., 152 (2022), 105766. https://doi.org/10.1016/j.ssci.2022.105766 doi: 10.1016/j.ssci.2022.105766
![]() |
[6] |
H. Gururaj, B. Swathi, R. Trupti, U. R. Darshan, A. Rajendra, K. Paramesha, Analysis of preventive measures against ddos attacks in smart grid, J. Inst. Eng. India, 104 (2023), 297–303. https://doi.org/10.1007/s40031-022-00844-1 doi: 10.1007/s40031-022-00844-1
![]() |
[7] |
A. K. Yadav, M. Misra, P. K. Pandey, M. Liyanage, An eap-based mutual authentication protocol for wlan-connected IoT devices, IEEE Trans. Ind. Inf., 19 (2022), 1343–1355. https://doi.org/10.1109/TII.2022.3194956 doi: 10.1109/TII.2022.3194956
![]() |
[8] |
R. Vinayakumar, M. Alazab, S. Srinivasan, Q. V. Pham, S. K. Padannayil, K. Simran, A visualized botnet detection system based deep learning for the internet of things networks of smart cities, IEEE Trans. Ind. Appl., 56 (2020), 4436–4456. https://doi.org/10.1109/TIA.2020.2971952 doi: 10.1109/TIA.2020.2971952
![]() |
[9] |
E. Benkhelifa, T. Welsh, W. Hamouda, A critical review of practices and challenges in intrusion detection systems for IoT: toward universal and resilient systems, IEEE Commun. Surv. Tut., 20 (2018), 3496–3509. https://doi.org/10.1109/COMST.2018.2844742 doi: 10.1109/COMST.2018.2844742
![]() |
[10] |
E. Bout, V. Loscri, A. Gallais, How machine learning changes the nature of cyberattacks on IoT networks: a survey, IEEE Commun. Surv. Tut., 24 (2021), 248–279. https://doi.org/10.1109/COMST.2021.3127267 doi: 10.1109/COMST.2021.3127267
![]() |
[11] |
A. Jamalipour, S. Murali, A taxonomy of machine-learning-based intrusion detection systems for the internet of things: a survey, IEEE Internet Things J., 9 (2021), 9444–9466. https://doi.org/10.1109/JIOT.2021.3126811 doi: 10.1109/JIOT.2021.3126811
![]() |
[12] |
N. W. Khan, M. S. Alshehri, M. A Khan, S. Almakdi, N. Moradpoor, A. Alazeb, et al., A hybrid deep learning-based intrusion detection system for IoT networks, Math. Biosci. Eng., 20 (2023), 13491–13520. https://doi.org/10.3934/mbe.2023602 doi: 10.3934/mbe.2023602
![]() |
[13] |
S. Ullah, J. Ahmad, M. A. Khan, M. S. Alshehri, W. Boulila, A. Koubaa, et al., TNN-IDS: transformer neural network-based intrusion detection system for MQTT-enabled IoT networks, Comput. Networks, 237 (2023), 110072. https://doi.org/10.1016/j.comnet.2023.110072 doi: 10.1016/j.comnet.2023.110072
![]() |
[14] |
S. Ullah, J. Ahmad, M. A. Khan, E. H. Alkhammash, M. Hadjouni, Y. Y. Ghadi, et al., A new intrusion detection system for the internet of things via deep convolutional neural network and feature engineering, Sensors, 22 (2022), 3607. https://doi.org/10.3390/s22103607 doi: 10.3390/s22103607
![]() |
[15] |
L. Cohen, G. Avrahami-Bakish, M. Last, A. Kandel, O. Kipersztok, Real-time data mining of non-stationary data streams from sensor networks, Inf. Fusion, 9 (2008), 344–353. https://doi.org/10.1016/j.inffus.2005.05.005 doi: 10.1016/j.inffus.2005.05.005
![]() |
[16] |
J. Lu, A. Liu, F. Dong, F. Gu, J. Gama, G. Zhang, Learning under concept drift: a review, IEEE Trans. Knowl. Data Eng., 31 (2018), 2346–2363. https://doi.org/10.1109/TKDE.2018.2876857 doi: 10.1109/TKDE.2018.2876857
![]() |
[17] |
J. Gama, P. Medas, G. Castillo, P. Rodrigues, Learning with drift detection, Brazilian Symposium on Artificial Intelligence, 2004,286–295. https://doi.org/10.1007/978-3-540-28645-5_29 doi: 10.1007/978-3-540-28645-5_29
![]() |
[18] |
A. Liu, J. Lu, F. Liu, G. Zhang, Accumulating regional density dissimilarity for concept drift detection in data streams, Pattern Recognit., 76 (2018), 256–272. https://doi.org/10.1016/j.patcog.2017.11.009 doi: 10.1016/j.patcog.2017.11.009
![]() |
[19] |
L. Du, Q. Song, X. Jia, Detecting concept drift: an information entropy based method using an adaptive sliding window, Intell. Data Anal., 18 (2014), 337–364. https://doi.org/10.3233/IDA-140645 doi: 10.3233/IDA-140645
![]() |
[20] |
H. Qiao, B. Novikov, J. O. Blech, Concept drift analysis by dynamic residual projection for effectively detecting botnet cyber-attacks in IoT scenarios, IEEE Trans. Ind. Inf., 18 (2021), 3692–3701. https://doi.org/10.1109/TII.2021.3108464 doi: 10.1109/TII.2021.3108464
![]() |
[21] |
L. Yang, A. Shami, A lightweight concept drift detection and adaptation framework for IoT data streams, IEEE Internet Things Mag., 4 (2021), 96–101. https://doi.org/10.1109/IOTM.0001.2100012 doi: 10.1109/IOTM.0001.2100012
![]() |
[22] |
W. Lalouani, M. Younis, Robust distributed intrusion detection system for edge of things, 2021 IEEE Global Communications Conference, 2021. https://doi.org/10.1109/GLOBECOM46510.2021.9685361 doi: 10.1109/GLOBECOM46510.2021.9685361
![]() |
[23] |
M. A. H. Al-Balhawi, G. Cansever, Intursion detection in iot networks using feature selection and svm classificastion, 2022 International Congress on Human-Computer Interaction, Optimization and Robotic Applications, 2022. https://doi.org/10.1109/HORA55278.2022.9799861 doi: 10.1109/HORA55278.2022.9799861
![]() |
[24] |
M. M. Alani, A. I. Awad, An intelligent two-layer intrusion detection system for the internet of things, IEEE Trans. Ind. Inf., 19 (2022), 683–692. https://doi.org/10.1109/TII.2022.3192035 doi: 10.1109/TII.2022.3192035
![]() |
[25] |
T. B. Hoang, L. Vu, Q. U. Nguyen, A data sampling and two-stage convolution neural network for IoT devices identification, 2022 RIVF International Conference on Computing and Communication Technologies, 2022,458–463. https://doi.org/10.1109/RIVF55975.2022.10013866 doi: 10.1109/RIVF55975.2022.10013866
![]() |
[26] |
G. Rathee, C. A. Kerrache, M. Lahby, Trustblksys: a trusted and blockchained cybersecure system for IIoT, IEEE Trans. Ind. Inf., 19 (2022), 1592–1599. https://doi.org/10.1109/TII.2022.3182984 doi: 10.1109/TII.2022.3182984
![]() |
[27] |
O. A. Wahab, Intrusion detection in the IoT under data and concept drifts: Online deep learning approach, IEEE Internet Things J., 9 (2022), 19706–19716. https://doi.org/10.1109/JIOT.2022.3167005 doi: 10.1109/JIOT.2022.3167005
![]() |
[28] |
E. Lughofer, E. Weigl, W. Heidl, C. Eitzinger, T. Radauer, Recognizing input space and target concept drifts in data streams with scarcely labeled and unlabelled instances, Inf. Sci., 355 (2016), 127–151. https://doi.org/10.1016/j.ins.2016.03.034 doi: 10.1016/j.ins.2016.03.034
![]() |
[29] |
C. Alippi, G. Boracchi, M. Roveri, Hierarchical change-detection tests, IEEE Trans. Neural Networks Learn. Syst., 28 (2016), 246–258. https://doi.org/10.1109/TNNLS.2015.2512714 doi: 10.1109/TNNLS.2015.2512714
![]() |
[30] |
K. Nishida, K. Yamauchi, Detecting concept drift using statistical testing, Proceedings of the 10th International Conference on Discovery Science, 2007,264–269. https://doi.org/10.1007/978-3-540-75488-6_27 doi: 10.1007/978-3-540-75488-6_27
![]() |
[31] |
R. S. M. de Barros, J. I. G. Hidalgo, D. R. de L. Cabral, Wilcoxon rank sum test drift detector, Neurocomputing, 275 (2018), 1954–1963. https://doi.org/10.1016/j.neucom.2017.10.051 doi: 10.1016/j.neucom.2017.10.051
![]() |
[32] |
L. Bu, C. Alippi, D. Zhao, A pdf-free change detection test based on density difference estimation, IEEE Trans. Neural Networks Learn. Syst., 29 (2016), 324–334. https://doi.org/10.1109/TNNLS.2016.2619909 doi: 10.1109/TNNLS.2016.2619909
![]() |
[33] |
D. Liu, Y. Wu, H. Jiang, FP-ELM: an online sequential learning algorithm for dealing with concept drift, Neurocomputing, 207 (2016), 322–334. https://doi.org/10.1016/j.neucom.2016.04.043 doi: 10.1016/j.neucom.2016.04.043
![]() |
[34] |
Y. Sun, K. Tang, Z. Zhu, X. Yao, Concept drift adaptation by exploiting historical knowledge, IEEE Trans. Neural Networks Learn. Syst., 29 (2018), 4822–4832. https://doi.org/10.1109/TNNLS.2017.2775225 doi: 10.1109/TNNLS.2017.2775225
![]() |
[35] |
D. S. Dimitrova, V. K. Kaishev, S. Tan, Computing the kolmogorov-smirnov distribution when the underlying cdf is purely discrete, mixed, or continuous, J. Stat. Software, 95 (2020), 1–42. https://doi.org/10.18637/jss.v095.i10 doi: 10.18637/jss.v095.i10
![]() |
[36] |
V. W. Berger, Y. Zhou, Kolmogorov-Smirnov test: overview, Wiley Statsref, 2014. https://doi.org/10.1002/9781118445112.stat06558 doi: 10.1002/9781118445112.stat06558
![]() |
[37] |
S. Yue, C. Wang, The Mann-Kendall test modified by effective sample size to detect trend in serially correlated hydrological series, Water Resour. Manage., 18 (2004), 201–218. https://doi.org/10.1023/B:WARM.0000043140.61082.60 doi: 10.1023/B:WARM.0000043140.61082.60
![]() |
[38] |
V. Hodge, J. Austin, A survey of outlier detection methodologies, Artif. Intell. Rev., 22 (2004), 85–126. https://doi.org/10.1023/B:AIRE.0000045502.10941.a9 doi: 10.1023/B:AIRE.0000045502.10941.a9
![]() |
[39] |
D. Olszewski, M. Iwanowski, W. Graniszewski, Dimensionality reduction for detection of anomalies in the IoT traffic data, Future Gener. Comput. Syst., 151 (2024), 137–151. https://doi.org/10.1016/j.future.2023.09.033 doi: 10.1016/j.future.2023.09.033
![]() |
[40] |
F. T. Liu, K. M. Ting, Z. H. Zhou, Isolation forest, 2008 Eighth IEEE International Conference on Data Mining, 2008,413–422. https://doi.org/10.1109/ICDM.2008.17 doi: 10.1109/ICDM.2008.17
![]() |
[41] |
F. T. Liu, K. M. Ting, Z. H. Zhou, Isolation-based anomaly detection, ACM Trans. Knowl. Discovery Data, 6 (2012), 1–39. https://doi.org/10.1145/2133360.2133363 doi: 10.1145/2133360.2133363
![]() |
[42] |
I. Souiden, M. N. Omri, Z. Brahmi, A survey of outlier detection in high dimensional data streams, Comput. Sci. Rev., 44 (2022), 100463. https://doi.org/10.1016/j.cosrev.2022.100463 doi: 10.1016/j.cosrev.2022.100463
![]() |
[43] |
L. Torlay, M. Perrone-Bertolotti, E. Thomas, M. Baciu, Machine learning-XGBoost analysis of language networks to classify patients with epilepsy, Brain Inf., 4 (2017), 159–169. https://doi.org/10.1007/s40708-017-0065-7 doi: 10.1007/s40708-017-0065-7
![]() |
[44] |
S. Mirjalili, A. Lewis, The whale optimization algorithm, Adv. Eng. Software, 95 (2016), 51–67. https://doi.org/10.1016/j.advengsoft.2016.01.008 doi: 10.1016/j.advengsoft.2016.01.008
![]() |
[45] |
F. S. Gharehchopogh, H. Gholizadeh, A comprehensive survey: whale optimization algorithm and its applications, Swarm Evol. Comput., 48 (2019), 1–24. https://doi.org/10.1016/j.swevo.2019.03.004 doi: 10.1016/j.swevo.2019.03.004
![]() |
[46] |
W. Yang, K. Xia, S. Fan, L. Wang, T. Li, J. Zhang, et al., A multi-strategy whale optimization algorithm and its application, Eng. Appl. Artif. Intell., 108 (2022), 104558. https://doi.org/10.1016/j.engappai.2021.104558 doi: 10.1016/j.engappai.2021.104558
![]() |
[47] |
W. N. Street, Y. Kim, A streaming ensemble algorithm (SEA) for large-scale classification, Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2001,377–382. https://doi.org/10.1145/502512.502568 doi: 10.1145/502512.502568
![]() |
[48] |
G. Widmer, M. Kubat, Learning in the presence of concept drift and hidden contexts, Mach. Learn., 23 (1996), 69–101. https://doi.org/10.1023/A:1018046501280 doi: 10.1023/A:1018046501280
![]() |
[49] |
G. Hulten, L. Spencer, P. Domingos, Mining time-changing data streams, Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2001, 97–106. https://doi.org/10.1145/502512.502529 doi: 10.1145/502512.502529
![]() |
[50] |
H. Wang, W. Fan, P. S. Yu, J. Han, Mining concept-drifting data streams using ensemble classifiers, Proceedings of the Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2003,226–235. https://doi.org/10.1145/956750.956778 doi: 10.1145/956750.956778
![]() |
[51] |
M. A. Ferrag, O. Friha, D. Hamouda, L. Maglaras, H. Janicke, Edge-IIoTset: a new comprehensive realistic cyber security dataset of IoT and IIoT applications for centralized and federated learning, IEEE Access, 10 (2022), 40281–40306. https://doi.org/10.1109/ACCESS.2022.3165809 doi: 10.1109/ACCESS.2022.3165809
![]() |
1. | Mohammed Shehu Shagari, Saima Rashid, Fahd Jarad, Mohamed S. Mohamed, Interpolative contractions and intuitionistic fuzzy set-valued maps with applications, 2022, 7, 2473-6988, 10744, 10.3934/math.2022600 | |
2. | Rehana Tabassum, Mohammed Shehu Shagari, Akbar Azam, OM Kalthum S. K. Mohamed, Awad A. Bakery, Muhammad Gulzar, Intuitionistic Fuzzy Fixed Point Theorems in Complex-Valued b -Metric Spaces with Applications to Fractional Differential Equations, 2022, 2022, 2314-8888, 1, 10.1155/2022/2261199 | |
3. | Xiaoming Qi, Zeeshan Ali, Tahir Mahmood, Peide Liu, Multi-Attribute Decision-Making Method Based on Complex Interval-Valued q-Rung Orthopair Linguistic Heronian Mean Operators and Their Application, 2023, 1562-2479, 10.1007/s40815-022-01455-0 | |
4. | Naveed Iqbal, Imran Khan, Rasool Shah, Kamsing Nonlaopon, The fuzzy fractional acoustic waves model in terms of the Caputo-Fabrizio operator, 2023, 8, 2473-6988, 1770, 10.3934/math.2023091 | |
5. | Wasfi Shatanawi, Taqi A. M. Shatnawi, Some fixed point results based on contractions of new types for extended b-metric spaces, 2023, 8, 2473-6988, 10929, 10.3934/math.2023554 | |
6. | Zeeshan Ali, Tahir Mahmood, Dragan Pamucar, Chuliang Wei, Complex Interval-Valued q-Rung Orthopair Fuzzy Hamy Mean Operators and Their Application in Decision-Making Strategy, 2022, 14, 2073-8994, 592, 10.3390/sym14030592 |