Class imbalance learning (CIL), which aims to addressing the performance degradation problem of traditional supervised learning algorithms in the scenarios of skewed data distribution, has become one of research hotspots in fields of machine learning, data mining, and artificial intelligence. As a postprocessing CIL technique, the decision threshold moving (DTM) has been verified to be an effective strategy to address class imbalance problem. However, no matter adopting random or optimal threshold designation ways, the classification hyperplane could be only moved parallelly, but fails to vary its orientation, thus its performance is restricted, especially on some complex and density variable data. To further improve the performance of the existing DTM strategies, we propose an improved algorithm called CDTM by dividing majority training instances into multiple different density regions, and further conducting DTM procedure on each region independently. Specifically, we adopt the well-known DBSCAN clustering algorithm to split training set as it could adapt density variation well. In context of support vector machine (SVM) and extreme learning machine (ELM), we respectively verified the effectiveness and superiority of the proposed CDTM algorithm. The experimental results on 40 benchmark class imbalance datasets indicate that the proposed CDTM algorithm is superior to several other state-of-the-art DTM algorithms in term of G-mean performance metric.
Citation: Mengke Lu, Shang Gao, Xibei Yang, Hualong Yu. Improving performance of decision threshold moving-based strategies by integrating density-based clustering technique[J]. Electronic Research Archive, 2023, 31(5): 2501-2518. doi: 10.3934/era.2023127
[1] | Cameron J. Browne, Chang-Yuan Cheng . Age-structured viral dynamics in a host with multiple compartments. Mathematical Biosciences and Engineering, 2020, 17(1): 538-574. doi: 10.3934/mbe.2020029 |
[2] | Andrey V. Melnik, Andrei Korobeinikov . Lyapunov functions and global stability for SIR and SEIR models withage-dependent susceptibility. Mathematical Biosciences and Engineering, 2013, 10(2): 369-378. doi: 10.3934/mbe.2013.10.369 |
[3] | Cameron Browne . Immune response in virus model structured by cell infection-age. Mathematical Biosciences and Engineering, 2016, 13(5): 887-909. doi: 10.3934/mbe.2016022 |
[4] | Ran Zhang, Shengqiang Liu . Global dynamics of an age-structured within-host viral infection model with cell-to-cell transmission and general humoral immunity response. Mathematical Biosciences and Engineering, 2020, 17(2): 1450-1478. doi: 10.3934/mbe.2020075 |
[5] | Jinliang Wang, Xiu Dong . Analysis of an HIV infection model incorporating latency age and infection age. Mathematical Biosciences and Engineering, 2018, 15(3): 569-594. doi: 10.3934/mbe.2018026 |
[6] | Andrey V. Melnik, Andrei Korobeinikov . Global asymptotic properties of staged models with multiple progression pathways for infectious diseases. Mathematical Biosciences and Engineering, 2011, 8(4): 1019-1034. doi: 10.3934/mbe.2011.8.1019 |
[7] | Jianxin Yang, Zhipeng Qiu, Xue-Zhi Li . Global stability of an age-structured cholera model. Mathematical Biosciences and Engineering, 2014, 11(3): 641-665. doi: 10.3934/mbe.2014.11.641 |
[8] | Jinhu Xu . Dynamic analysis of a cytokine-enhanced viral infection model with infection age. Mathematical Biosciences and Engineering, 2023, 20(5): 8666-8684. doi: 10.3934/mbe.2023380 |
[9] | Xichao Duan, Sanling Yuan, Kaifa Wang . Dynamics of a diffusive age-structured HBV model with saturating incidence. Mathematical Biosciences and Engineering, 2016, 13(5): 935-968. doi: 10.3934/mbe.2016024 |
[10] | Abdennasser Chekroun, Mohammed Nor Frioui, Toshikazu Kuniya, Tarik Mohammed Touaoula . Global stability of an age-structured epidemic model with general Lyapunov functional. Mathematical Biosciences and Engineering, 2019, 16(3): 1525-1553. doi: 10.3934/mbe.2019073 |
Class imbalance learning (CIL), which aims to addressing the performance degradation problem of traditional supervised learning algorithms in the scenarios of skewed data distribution, has become one of research hotspots in fields of machine learning, data mining, and artificial intelligence. As a postprocessing CIL technique, the decision threshold moving (DTM) has been verified to be an effective strategy to address class imbalance problem. However, no matter adopting random or optimal threshold designation ways, the classification hyperplane could be only moved parallelly, but fails to vary its orientation, thus its performance is restricted, especially on some complex and density variable data. To further improve the performance of the existing DTM strategies, we propose an improved algorithm called CDTM by dividing majority training instances into multiple different density regions, and further conducting DTM procedure on each region independently. Specifically, we adopt the well-known DBSCAN clustering algorithm to split training set as it could adapt density variation well. In context of support vector machine (SVM) and extreme learning machine (ELM), we respectively verified the effectiveness and superiority of the proposed CDTM algorithm. The experimental results on 40 benchmark class imbalance datasets indicate that the proposed CDTM algorithm is superior to several other state-of-the-art DTM algorithms in term of G-mean performance metric.
In the early 1930s, in order to generalize the formula of quantum mechanics, Jordan et al. introduced an important commutative non-associative algebra [1], which was initially called "r-order digital system". In 1947, Albert renamed this kind of algebra Jordan algebra and studied their structural theory [2]. Since then, Jordan algebras have attracted extensive attention. Particularly, Jacobson developed the representation theory of Jordan algebras [3,4]. Jordan superalgebras were first studied by Kac, who classified simple finite dimensional Jordan superalgebras over an algebraically closed field of characteristic zero [5]. Jordan superalgebras also have significant applications in quantum mechanics [6,7]. More results on Jordan superalgebras are available in [8,9].
Hom-type algebras were first introduced to study the q-deformation of Witt and Virasoro algebras [10,11], which played an important role in physics, mainly in conformal field theory. Bihom-type algebras are generalizations of Hom-type algebras, which were presented by Graziani et al. from the categorical point and applied to study certain deformations of quantum groups [12]. Up to now, the (Bi)hom-structures of various algebras have been intensively investigated. The construction relationship between Hom-type algebras and the module structure on them can be found in the literature [13,14,15,16,17]. Naturally, the construction between Bihom-type algebras is studied in the literature [18,19], and the results of representation and deformation can be found in [20,21,22]. In this paper, we first generalize bimodules and representations of Bihom-Jordan algebras [23,24] to Bihom-Jordan superalgebras and then develop the theory of representations and O-operators on Bihom-Jordan superalgebras.
The outline of the paper is presented as follows: In Section 2, we review some basics about Bihom-superalgebras, Bihom-Jordan superalgebras; we study Bihom-super modules and give some easy constructions of Bihom-Jordan superalgebras. In Section 3, we mainly study super-bimodules on Bihom-Jordan superalgebras and obtain some new constructions under the view of module. In Section 4, we study the representation of Bihom-Jordan superalgebra and give the definitions of O-operator and Rota–Baxter operator. At the same time, we also give the definition of Bihom-pre-Jordan superalgebra. Finally, the relationship between O-operator and Bihom-pre-Jordan superalgebra is studied. Actually, on the basis of this section, we can also continue to study cohomology theory.
Throughout the paper, all algebraic systems are supposed to be over a field of characteristic 0. Let A be a linear superspace over K that is a Z2-graded linear space with a direct sum A=A¯0⊕A¯1. The elements of Aj,j=0,1, are said to be homogenous and of parity j. The parity of a homogeneous element x is denoted by |x|. In the sequel, we will denote by H(A) the set of all homogeneous elements of A. In this paper, we need to use the elements, all of which are not specified, are homogeneous.
In this section, we recall some basic definitions about Bihom-Jordan superalgebras, provide some construction results. A Bihom-superalgebra is a quadruple (J,μ,α,β), where μ:J⊗J→J is an even bilinear map and α,β:J→J are even linear maps such that α∘μ=μ∘α and β∘μ=μ∘β (multiplicativity).
Definition 2.1. [25] Let (J,μ,α,β) be a Bihom-superalgebra.
The Bihom-associator of J is an even trilinear map asα,β:J⊗3⟶J defined by
asα,β=μ∘(μ⊗β−α⊗μ). | (2.1) |
For any ε,γ,δ∈H(J), asα,β(ε,γ,δ)=μ(μ(ε,γ),β(δ))−μ(α(ε),μ(γ,δ)).
In particular, when α=β=Id, Bihom-superalgebra is to degenerate to the superalgebra, so is Bihom-associator degenerates to the original associator. If α=β, Bihom-associator degenerates to the Hom-associator.
Definition 2.2. Let (J,μ,α,β) be a Bihom-superalgebra. Then
● A Bihom-sub-superalgebra of J is a Z2-graded linear subspace B⊆J, which satisfies μ(ε,γ)∈B,α(ε)∈B and β(ε)∈B, for all ε,γ∈H(J). Furthermore, if μ(ε,γ),μ(γ,ε)∈B, for all (ε,γ)∈J×B, then B is called a two-sided Bihom-ideal of J.
● J is regular if α and β are algebra automorphisms.
● J is involutive if α and β are two involutions, that is α2=β2=Id.
Definition 2.3. Let (J,μ,α,β) and (J′,μ′,α′β′) be two Bihom-superalgebras. If a homomorphism f:J→J′ satisfies the following conditions:
f∘μ=μ′∘(f⊗f),f∘α=α′∘fandf∘β=β′∘f. |
Then f is called Bihom-superalgebra morphism. And we call the set Γf={ε+f(ε)|ε∈H(J)}⊂J⊕J′ the graph of f.
Proposition 2.1. Let (J,μJ,αJ,βJ) and (B,μB,αB,βB) be two Bihom-Jordan superalgebras. Then an even linear map f:J→B is a morphism if and only if its graph Γf is a Bihom-subalgebra of (J⊕B,μ=μJ+μB,α=αJ+αB,β=βJ+βB).
Proof. Suppose that f is a morphism of Bihom-Jordan superalgebras. Clearly, Γf is a subspace of J⊕B, we only need to prove the Γf is closed under the μ,α,β. For all ε,γ∈H(J),
μ(ε+f(ε),γ+f(γ))=μJ(ε,γ)+μB(f(ε),f(γ))=μJ(ε,γ)+fμJ(ε,γ). |
Moreover, by fαJ=αBf and fβJ=βBf,
α(ε+f(ε))=αJ(ε)+αB(f(ε))=αJ(ε)+fαJ(ε),β(ε+f(ε))=βJ(ε)+βB(f(ε))=βJ(ε)+fβJ(ε). |
It follows that Γf is a Bihom-subalgebra of J⊕B.
Conversely, Γf is a Bihom-subalgebra of J⊕B, so
μ(ε+f(ε),γ+f(γ))=μJ(ε,γ)+μB(f(ε),f(γ))∈Γf, |
which implies that μB(f(ε),f(γ))=fμJ(ε,γ). Similarly, we also obtain αBf=fαJ and βBf=fβJ from α(Γf)⊆Γf and β(Γf)⊆Γf, respectively. Thus, f is a morphism of Bihom-Jordan superalgebras.
Definition 2.4. [25] A Bihom-associative superalgebra is a quadruple (J,μ,α,β), where α,β:J→J are even linear maps and μ:J×J→J is an even bilinear map such that αβ=βα,αμ=μα⊗2, βμ=μβ⊗2 and satisfying Bihom-associator is zero:
asα,β(ε,γ,δ)=0,forallε,γ,δ∈H(J). (Bihom-associativity condition) |
Clearly, when α=β, we obtain a Hom-associative superalgebra.
Definition 2.5. [19] A BiHom superalgebra (J,μ,α,β) is called a Bihom-Jordan superalgebra if for all ε,γ,δ,t∈H(J):
(i) αβ=βα,(ii) μ(β(ε),α(γ))=(−1)|ε||γ|μ(β(γ),α(ε)),(Bihom-super commutativity condition)(iii) ↺ε,γ,t(−1)|t|(|ε|+|δ|)~asα,β(μ(β2(ε),αβ(γ)),α2β(δ),α3(t))=0.(Bihom-Jordan super-identity) |
In particular, it is reduced to a Jordan superalgebra when α=β=Id.
Next, we give some common construction methods. Let (J,μ,α,β) be a Bihom-superalgebra. Define its plus Bihom-superalgebra as the Bihom-superalgebra J+=(J,∗,α,β), where
ε∗γ=12(μ(ε,γ)+(−1)|ε||γ|μ(α−1β(γ),αβ−1(ε))). |
Note that product ∗ is Bihom-supercommutative. In fact, for all ε,γ∈H(J),
β(ε)∗α(γ)=12(β(ε)α(γ)+(−1)|ε||γ|β(γ)α(ε))=(−1)|ε||γ|12(β(γ)α(ε)+β(ε)α(γ))=(−1)|ε||γ|β(γ)∗α(ε). |
Moreover, the plus Bihom-superalgebra J+=(J,∗,α,β) is a Bihom-Jordan superalgebra. Naturally, we define
ε⋄γ=μ(ε,γ)+(−1)|ε||γ|μ(α−1β(γ),αβ−1(ε)), |
the ⋄ is also Bihom-supercommutative. Then J⋄=(J,⋄,α,β) is also a Bihom-Jordan superlagebra.
Besides that, Bihom-Jordan superalgebra (J,μα,β=μ(α⊗β),α,β) can be obtained from Jordan superalgebra (J,μ). We also consider the quotient algebra obtained by modulo Bihom-ideal, given a Bihom-Jordan superalgebra (J,μ,α,β) and I is a Bihom-ideal. Define ˉμ,ˉα,ˉβ on J/I as follows:
ˉμ(ˉε,ˉγ)=¯μ(ε,γ),ˉα(ˉε)=¯α(ε),ˉβ(ˉε)=¯β(ε). |
Then (J/I,ˉμ,ˉα,ˉβ) is also a Bihom-Jordan superalgebra.
Example 2.1. Given a 3-dimensional Jordan superalgebra (J=Jˉ0⊕Jˉ1,μ) in [5], the bases of Jˉ0 and Jˉ1 are {ε} and {u,v}, respectively. The nontrivial multiplication is defined as follows:
μ(ε,ε)=ε,μ(ε,u)=12u,μ(ε,v)=12v,μ(u,v)=ε. |
We consider two even endomorphisms α and β, which satisfy α(ε)=ε,α(u)=−u,α(v)=−v, and β(ε)=−ε,β(u)=−u,β(v)=v. Then we obtain a Bihom-Jordan superalgebra (J,μ′=μ(α⊗β),α,β).
Example 2.2. In [5], let (J=Jˉ0⊕Jˉ1,μ) be a Jordan superalgebra with the nontrivial multiplication as follows:
μ(ε,ε)=2ε,μ(ε,u)=u,μ(ε,v)=v,μ(u,v)=1+kx, |
k∈K and k≠12, where {1,ε} and {u,v} are bases of Jˉ0 and Jˉ1, respectively. We define two even endomorphisms α and β satisfies α(1)=1,α(ε)=ε,α(u)=−u,α(v)=−v and β(1)=1,β(ε)=−ε,β(u)=u,β(v)=v. Then we obtain a Bihom-Jordan superalgebra (J,μ′=μ(α⊗β),α,β).
Definition 2.6. Let (J,μ,α,β) be a Bihom-superalgebra.
1) A Bihom-super-module (V,ϕ,ψ) is called an J-super-bimodule if it is equipped with an even left structure ρl and an even right structure map ρr on Z2-graded vector space V, ρl and ρr are given by
● ρl:(J⊗V,α⊗ϕ,β⊗ψ)→(V,ϕ,ψ),ρl(a,v)=a⋅v,
● ρr:(V⊗J,ϕ⊗α,ψ⊗β)→(V,ϕ,ψ),ρr(v,a)=v⋅a.
2) An even linear map f:(V,ϕ,ψ,ρl,ρr)→(V′,ϕ′,ψ′,ρ′l,ρ′r) is a morphism of the Bihom-super-modules such that the following commutative diagrams
![]() |
3) Let (V,ϕ,ψ,ρl,ρr) be an J-super-bimodule. Then the module Bihom-associator asVϕ,ψ of V is defined as:
asVϕ,ψ∘IdV⊗J⊗J=ρr∘(ρr⊗β)−ρr∘(ϕ⊗μ), | (2.2) |
asVϕ,ψ∘IdJ⊗V⊗J=ρr∘(ρl⊗β)−ρl∘(α⊗ρr), | (2.3) |
asVϕ,ψ∘IdJ⊗J⊗V=ρl∘(μ⊗ψ)−ρl∘(α⊗ρl). | (2.4) |
Definition 2.7. Let (J,μ,α,β) be a Bihom-associative superalgebra and (V,ϕ,ψ) be a Bihom-super-module. Then
1) A left Bihom-associative J-super-module structure consists of an even morphism ρl:J⊗V→V satisfies asVϕ,ψ=0 in (2.4).
2) A right Bihom-associative J-super-module structure consists of an even morphism ρr:V⊗J→V satisfies asVϕ,ψ=0 in (2.2).
3) A Bihom-associative J-super-bimodule structure consists of an even morphism ρl:J⊗V→V and an even morphism ρr:V⊗J→V such that (V,ϕ,ψ,ρl) is a left Bihom-associative J-super-module, (V,ϕ,ψ,ρr) is a right Bihom-associative J-super-module, and satisfies asVϕ,ψ=0 in (2.3).
In this section, we introduce super-bimodules of Bihom-Jordan superalgebras and give some of their constructions. Finally, we define an abelian extension in order to give an application in the next section. For convenience, the sign will subsequently be omitted from the product operation of elements in J.
Definition 3.1. Let (J,μ,α,β) be a Bihom-Jordan superalgebra. For all ε,γ,δ∈H(J),v∈H(V),
● A left Bihom-Jordan J-super-module is a Bihom-super-module (V,ϕ,ψ) that is equipped with an even left structure map ρl:J⊗V→V,ρl(a⊗v)=a⋅v such that ψ is invertible and the following conditions hold:
↺ε,γ,δ(−1)|γ||δ|β2α2(δ)⋅(αβ(ε)α2(γ)⋅ϕ3(v))=↺ε,γ,δ(−1)|ε||δ|αβ2(ε)α2β(γ)⋅(βα2(δ)⋅ϕ3(v)), | (3.1) |
β2α2(δ)⋅(βα2(γ)⋅(α2(ε)⋅ψ−1ϕ3(v)))+(−1)|ε||γ|+|ε||δ|+|γ||δ|β2α2(ε)⋅(βα2(γ)⋅(α2(δ)⋅ψ−1ϕ3(v)))+(−1)|ε||δ|+|ε||γ|((β2(ε)βα(δ))βα2(γ))⋅ϕ3ψ(v)=(−1)|γ||δ|β2α(γ)βα2(δ)⋅(βα2(ε)⋅ϕ3(v))+(−1)|γ||δ|+|ε||δ|β2α(γ)βα2(ε)⋅(βα2(δ)⋅ϕ3(v))+(−1)|ε||δ|+|ε||γ|β2α(ε)βα2(δ)⋅(βα2(γ)⋅ϕ3(v)). | (3.2) |
● A right Bihom-Jordan J-super-module is a Bihom-super-module (V,ϕ,ψ) that is equipped with an even right structure map ρr:V⊗J→V,ρr(v⊗a)=v⋅a such that the following conditions hold:
↺ε,γ,δ(−1)|ε||δ|(ϕψ2(v)⋅αβ(ε)α2(γ))⋅βα3(δ)=↺ε,γ,δ(−1)|γ||δ|(ϕψ2(v)⋅βα2(δ))⋅α2β(ε)α3(γ). | (3.3) |
((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|ε||γ|+|ε||δ|+|γ||δ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)+(−1)|γ||δ|ϕ2ψ2(v)⋅(βα(ε)α2(δ))α3(γ)=(ϕψ2(v)⋅βα2(ε))⋅α2β(γ)α3(δ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|(ϕψ2(v)⋅βα2(δ))⋅α2β(γ)α3(ε)+(−1)|ε||γ|(ϕψ2(v)⋅βα2(γ))⋅α2β(ε)α3(δ). | (3.4) |
Theorem 3.1. Let (J,μ,α,β) be a Bihom-Jordan superalgebra, (V,ϕ,ψ) be a Bihom-super-module and ρr:V⊗J→V,ρr(v⊗a)=v⋅a be an even linear map, which satisfies the following conditions: for all ε,γ∈H(J),v∈H(V),
ϕρr=ρr(ϕ⊗α),ψρr=ρr(ψ⊗β), | (3.5) |
ϕ(v)⋅β(ε)α(γ)=(v⋅β(ε))⋅βα(γ)+(−1)|ε||γ|(v⋅β(γ))⋅αβ(ε). | (3.6) |
Then (V,ϕ,ψ,ρr) is a right Bihom-Jordan J-super-module, called a right special Bihom-Jordan J-super-module.
Proof. For any ε,γ,δ∈H(J),v∈H(V),
↺ε,γ,δ(−1)|δ||γ|(ϕψ2(v)⋅βα2(δ))⋅α2β(ε)α3(γ)=↺ε,γ,δ(−1)|δ||γ|ϕ(ψ2(v)⋅βα(δ))⋅α2β(ε)α3(γ)(by(3.5))=↺ε,γ,δ(−1)|δ||γ|((ψ2(v)⋅βα(δ))⋅βα2(ε))⋅βα3(γ)+↺ε,γ,δ(−1)|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)(by(3.6))=↺ε,γ,δ(−1)|δ||γ|(ϕψ2(v)⋅βα(δ)α2(ε))⋅βα2(γ)−↺ε,γ,δ(−1)|δ||γ|+|δ||ε|((ψ2(v)⋅βα(ε))⋅βα2(δ))⋅βα3(γ)+↺ε,γ,δ(−1)|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)(by(3.6))=↺ε,γ,δ(−1)|ε||δ|(ϕψ2(v)⋅βα(ε)α2(γ))⋅βα2(δ). |
So Eq (3.3) holds. On the other hand,
(ϕψ2(v)⋅βα2(ε))⋅βα2(γ)α3(δ)+(−1)|ε||γ|(ϕψ2(v)⋅βα2(γ))⋅βα2(ε)α3(δ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|(ϕψ2(v)⋅βα2(δ))⋅βα2(γ)α3(ε)=ϕ(ψ2(v)⋅βα(ε))⋅βα2(γ)α3(δ)+(−1)|ε||γ|ϕ(ψ2(v)⋅βα(γ))⋅βα2(ε)α3(δ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|ϕ(ψ2(v)⋅βα(δ))⋅βα2(γ)α3(ε)(by(3.5))=((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|γ||δ|((ψ2(v)⋅βα(ε))⋅βα2(δ))⋅βα3(γ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)+(−1)|δ||ε|+|δ||γ|((ψ2(v)⋅βα(δ))⋅βα2(ε))⋅βα3(γ)+(−1)|ε||γ|((ψ2(v)⋅βα(γ))⋅βα2(ε))⋅βα3(δ)+(−1)|ε||γ|+|ε||δ|((ψ2(v)⋅βα(γ))⋅βα2(δ))⋅βα3(ε)(by(3.6))=((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|γ||δ|((ϕψ2(v)⋅βα(ε)α2(δ))⋅α3β(γ)−(−1)|γ||δ|+|ε||δ|((ψ2(v)⋅βα(δ))⋅βα2(ε))⋅βα3(γ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)+(−1)|δ||ε|+|δ||γ|((ψ2(v)⋅βα(δ))⋅βα2(ε))⋅βα3(γ)+(−1)|ε||γ|((ψ2(v)⋅βα(γ))⋅βα2(ε))⋅βα3(δ)+(−1)|ε||γ|+|ε||δ|((ψ2(v)⋅βα(γ))⋅βα2(δ))⋅βα3(ε)(by(3.6))=((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|γ||δ|((ϕψ2(v)⋅βα(ε)α2(δ))⋅α3β(γ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)+(−1)|ε||γ|((ψ2(v)⋅βα(γ))⋅βα2(ε))⋅βα3(δ)+(−1)|ε||γ|+|ε||δ|((ψ2(v)⋅βα(γ))⋅βα2(δ))⋅βα3(ε)=((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|γ||δ|ϕ2ψ2(v)⋅(βα(ε)α2(δ))α3(δ)−(−1)|ε||γ|(ϕψ2(v)⋅βα2(γ))⋅βα2(ε)α3(δ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)+(−1)|ε||γ|((ψ2(v)⋅βα(γ))⋅βα2(ε))⋅βα3(δ)+(−1)|ε||γ|+|ε||δ|((ψ2(v)⋅βα(γ))⋅βα2(δ))⋅βα3(ε)(by(3.6))=((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅βα3(δ)+(−1)|γ||δ|ϕ2ψ2(v)⋅(βα(ε)α2(δ))α3(δ)+(−1)|δ||ε|+|δ||γ|+|ε||γ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅βα3(ε)(by(3.6)). |
It follows Eq (3.4).
Similarly, we have the following result.
Theorem 3.2. Let (J,μ,α,β) be a Bihom-Jordan superalgebra, (V,ϕ,ψ) be a Bihom-super-module such that ψ is invertible, and ρl:J⊗V→V be an even linear map given by ρl(a⊗v)=a⋅v such that the following conditions hold:
ϕρl=ρl(α⊗ϕ),ψρl=ρl(β⊗ψ), | (3.7) |
β(ε)α(γ)⋅ψ(v)=βα(ε)⋅(α(γ)⋅v)+(−1)|ε||γ|βα(γ)⋅(α(ε)⋅v). | (3.8) |
Then (V,ϕ,,ψ,ρl) is a left Bihom-Jordan J-super-module called a left special super-module.
Proof. Similar to the proof of Theorem 3.1, the conclusion can be proved by repeatedly using Eqs (3.7) and (3.8).
Now, we give the definition super-bimodule of a BiHom-Jordan superalgebra.
Definition 3.2. Let (J,μ,α,β) be a Bihom-Jordan superalgebra. A Bihom-Jordan J-super-bimodule is a Bihom-super-module (V,ϕ,ψ) with an even left structure map ρl:J⊗V→V,ρl(a⊗v)=a⋅v and an even right structure map ρr:V⊗J→V,ρr(v⊗a)=v⋅a satisfying three conditions:
ρl(β⊗ϕ)=ρr(ψ⊗α)τ1, | (3.9) |
↺ε,γ,δ(−1)|δ|(|ε|+|v|)asVϕ,ψ(μ(β2(ε),αβ(γ)),ϕ2ψ(v),α3(δ))=0, | (3.10) |
(−1)|γ||δ|asVϕ,ψ(ψ2(v)⋅αβ(ε),βα2(γ),α3(δ))+(−1)|ε||γ|+|ε||δ|asVϕ,ψ(ψ2(v)⋅αβ(δ),βα2(γ),α3(ε))+(−1)|v||ε|+|v||γ|+|v||δ|asVϕ,ψ(μ(β2(ε),αβ(δ)),βα2(γ),ϕ3(v))=0. | (3.11) |
Remark 3.1. 1) If α=β=IdJ and ϕ=ψ=IdV then V is reduced to the so-called Jordan supermodule of the Jordan superalgebra (J,μ).
2) Clearly, a Bihom-Jordan A-super-bimodule is a right Bihom-Jordan super-module. Furthermore, it is a left Bihom-Jordan super-module if ψ is invertible.
Example 3.1. Here are some examples of Bihom-Jordan super-bimodules.
1) Let (J,μ,α,β) be a Bihom-Jordan superalgebra. Then (J,α,β) is a Bihom-Jordan J-super-bimodule where the structure maps are ρl=ρr=μ. More generally, if B is a Bihom-ideal of (J,μ,α,β), then (B,α,β) is a Bihom-Jordan J-super-bimodule where the structure maps are ρl(a,ε)=μ(a,ε)=μ(ε,a)=ρr(ε,a), for all (a,ε)∈H(J)×H(B).
2) If (J,μ) is a Jordan superalgebra and M is a Jordan J-super-bimodule in the usual sense, then (M,IdM,IdM) is a BiHom-Jordan J-super-bimodule where (J,μ,IdJ,IdJ) is a Bihom-Jordan superalgebra.
Theorem 3.3. Let (J,μ,α,β) be a Bihom-Jordan superalgebra and (V,ϕ,ψ,ρl,ρr) be a Bihom-Jordan J-super-bimodule. Define even linear maps ˜μ,˜α and ˜β on J⊕V,
● ˜μ:(J⊕V)⊗2→J⊕V,˜μ(ε+u,γ+v):=μ(ε,γ)+ε⋅v+u⋅γ,
● ˜α,˜β:(J⊕V)→J⊕V,
˜α(ε+u):=α(ε)+ϕ(v) and ˜β(ε+u):=β(ε)+ψ(v).
Then (J⊕V,˜μ,˜α,˜β) is a Bihom-Jordan superalgebra.
Proof. We omitted the calculation process; it is straightforward to see Bihom-super commutativity condition and Bihom-Jordan super-identity by Definition 3.2.
The next result shows that a special left and right Bihom-Jordan super-module has a Bihom-Jordan super-bimodule structure under a specific condition.
Theorem 3.4. Let (J,μ,α,β) be a regular Bihom-Jordan superalgebra, (V,ϕ,ψ) be both a left and a right special BiHom-Jordan J-module with the structure maps ρ1 and ρ2 respectively, such that ϕ is invertible, and the Bihom-associativity condition holds
ρ2∘(ρ1⊗β)=ρ1∘(α⊗ρ2). | (3.12) |
Define two even bilinear maps ρl:J⊗V→V and ρr:V⊗J→V by
ρl=ρ1+ρ2(ψϕ−1⊗αβ−1)∘τ1andρr=ρ1(βα−1⊗ϕψ−1)∘τ2+ρ2. | (3.13) |
Then (V,ϕ,ψ,ρl,ρr) is a Bihom-Jordan J-super-bimodule.
Proof. ρl and ρr are even structure maps from ρ1 and ρ2. We need to check out (3.9)–(3.11). First, for any (ε,v)∈H(J)×H(V),
ρl(β(ε),ϕ(v))=β(ε)⋅ϕ(v)+(−1)|a||v|ψϕ−1(ϕ(v))⋅αβ−1(β(ε))=β(ε)⋅ϕ(v)+(−1)|a||v|ψ(v)⋅α(ε), |
ρr(ψ⊗α)τ1(ε⊗v)=(−1)|a||v|ρr(ψ(v),α(ε))=(−1)|a||v|ψ(v)⋅α(ε)+βα−1(α(ε))⋅ϕψ−1(ϕ(v))=β(ε)⋅ϕ(v)+(−1)|a||v|ψ(v)⋅α(ε). |
So ρl(β⊗ϕ)=ρr(ψ⊗α)τ1. Next, for any ε,γ,δ∈H(J),v∈H(V)
asVϕ,ψ(μ(β2(ε),αβ(γ)),ϕ2ψ(v),α3(δ))=ρr(ρl(μ(β2(ε),αβ(γ)),ϕ2ψ(v)),βα3(δ))−ρl(αβ2(ε)α2β(γ),ρr(ϕ2ψ(v),α3(δ)))=ρr(β2(ε)αβ(γ)⋅ϕ2ψ(v),βα3(δ))+(−1)|ε||v|+|γ||v|ρr(ϕψ2(v)⋅αβ(ε)α2(γ),βα3(δ))−(−1)|v||δ|ρl(αβ2(ε)α2β(γ),βα2(δ)⋅ϕ3(v))−ρl(αβ2(ε)α2β(γ),ϕ2ψ(v)⋅α3(δ))(by(3.13))=(β2(ε)αβ(γ)⋅ϕ2ψ(v))⋅βα3(δ)+(−1)|δ||ε|+|δ||γ|+|δ||v|α2β2(δ)⋅(αβ(ε)α2(γ)⋅ϕ3(v))+(−1)|ε||v|+|γ||v|(ϕψ2(v)⋅αβ(ε)α2(γ))⋅βα3(δ)+(−1)|ε||v|+|γ||v|+|δ||v|+|δ||ε|+|δ||γ|α2β2(δ)⋅(ϕ2ψ(v)⋅α2(ε)α3β−1(γ))−(−1)|v||δ|αβ2(ε)α2β(γ)⋅(βα2(δ)⋅ϕ3(v))−(−1)|v||δ|+|δ||ε|+|δ||γ|+|v||ε|+|v||γ|(β2α(δ)⋅ψϕ2(v))⋅α2β(ε)α3(γ)−αβ2(ε)α2β(γ)⋅(ϕ2ψ(v)⋅α3(δ))−(−1)|v||ε|+|v||γ|+|δ||ε|+|δ||γ|(ϕψ2(v)⋅α2β(δ))⋅α2β(ε)α3(γ)(by(3.13))=(−1)|ε||v|+|γ||v|(ϕψ2(v)⋅αβ(ε)α2(γ))⋅βα3(δ)−(−1)|v||ε|+|v||γ|+|δ||ε|+|δ||γ|(ϕψ2(v)⋅α2β(δ))⋅α2β(ε)α3(γ)+(−1)|δ||ε|+|δ||γ|+|δ||v|α2β2(δ)⋅(αβ(ε)α2(γ)⋅ϕ3(v))−(−1)|v||δ|αβ2(ε)α2β(γ)⋅(βα2(δ)⋅ϕ3(v))(by(3.12)). |
So
↺ε,γ,δ(−1)|δ|(|ε|+|v|)asVϕ,ψ(β2(ε)αβ(γ),ϕ2ψ(v),α3(δ))=(−1)|v|(|ε|+|γ|+|δ|){↺ε,γ,δ(−1)|ε||δ|(ϕψ2(v)⋅αβ(ε)α2(γ))⋅βα3(δ)−↺ε,γ,δ(−1)|δ||γ|(ϕψ2(v)⋅βα2(δ))⋅α2β(ε)α3(γ)}+↺ε,γ,δ(−1)|δ||γ|α2β2(δ)⋅(αβ(ε)α2(γ)⋅ϕ3(v))−↺ε,γ,δ(−1)|δ||ε|αβ2(ε)α2β(γ)⋅(βα2(δ)⋅ϕ3(v))=(−1)|v|(|ε|+|γ|+|δ|)0+0=0. |
Finally, to prove (3.11), let us compute each of its three terms.
(−1)|γ||δ|asVϕ,ψ(ρr(ψ2(v),βα(ε)),βα2(γ),α3(δ))=(−1)|γ||δ|asVϕ,ψ(ψ2(v)⋅βα(ε),βα2(γ),α3(δ))+(−1)|γ||δ|+|ε||v|asVϕ,ψ(β2(ε)⋅ϕψ(v),βα2(γ),α3(δ))(by(3.13))=(−1)|γ||δ|ρr(ρr(ψ2(v)⋅βα(ε),βα2(γ)),α3β(δ))−(−1)|γ||δ|ρr(ϕψ2(v)⋅βα2(ε),βα2(γ)α3(δ))+(−1)|γ||δ|+|ε||v|ρr(ρr(β2(ε)⋅ϕψ(v),βα2(γ)),α3β(δ))−(−1)|γ||δ|+|ε||v|ρr(αβ2(ε)⋅ϕ2ψ(v),βα2(γ)α3(δ))=(−1)|γ||δ|ρr((ψ2(v)⋅βα(ε))⋅βα2(γ),α3β(δ))+(−1)|γ||δ|+|γ||v|+|ε||γ|ρr(β2α(γ)⋅(ϕψ(v)⋅α2(ε)),α3β(δ)−(−1)|γ||δ|(ϕψ2(v)⋅βα2(ε))⋅βα2(γ)α3(δ)−(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|β2α(γ)βα2(δ)⋅(ϕ2ψ(v)⋅α3(ε))+(−1)|γ||δ|+|ε||v|ρr((β2(ε)⋅ϕψ(v))⋅βα2(γ),α3β(δ))+(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|ρr(β2α(γ)⋅(αβ(ε)⋅ϕ2(v)),α3β(δ))−(−1)|γ||δ|+|ε||v|(αβ2(ε)⋅ϕ2ψ(v))⋅βα2(γ)α3(δ)−(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|β2α(γ)βα2(δ)⋅(α2β(γ)⋅ϕ3(v))(by(3.13))=(−1)|γ||δ|((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅α3β(δ)+(−1)|δ||ε|+|δ||v|α2β2(δ)⋅((ϕψ(v)⋅α2(ε))⋅α3(γ))+(−1)|γ||δ|+|γ||v|+|ε||γ|(β2α(γ)⋅(ϕψ(v)⋅α2(ε)))⋅α3β(δ)+(−1)|γ||v|+|ε||γ|+|δ||v|+|δ||ε|α2β2(δ)⋅(βα2(γ)⋅(ϕ2(v)⋅α3β−1(ε)))+(−1)|γ||δ|+|ε||v|((β2(ε)⋅ϕψ(v))⋅βα2(γ))⋅α3β(δ)+(−1)|ε||v|+|δ||ε|+|δ||v|α2β2(δ)⋅((αβ(ε)⋅ϕ2(v))α3(γ))+(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|(β2α(γ)⋅(αβ(ε)⋅ϕ2(v)))⋅α3β(δ)+(−1)|ε||v|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|α2β2(δ)⋅(α2β(γ)⋅(α2(ε)⋅ϕ3ψ−1(v)))−(−1)|γ||δ|(ϕψ2(v)⋅βα2(ε))⋅βα2(γ)α3(δ)−(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|β2α(γ)βα2(δ)⋅(ϕ2ψ(v)⋅α3(ε))⏟B−(−1)|γ||δ|+|ε||v|(αβ2(ε)⋅ϕ2ψ(v))⋅βα2(γ)α3(δ)⏟J−(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|β2α(γ)βα2(δ)⋅(α2β(γ)⋅ϕ3(v)).(by(3.13)andrearranging) |
Observe that
J=(−1)|γ||δ|+|ε||v|ϕ(β2(ε)⋅ϕψ(v))⋅β(α2(γ))α(α2(δ))=(−1)|γ||δ|+|ε||v|((β2(ε)⋅ϕψ(v))⋅βα2(γ))⋅βα3(δ)+(−1)|ε||v|((β2(ε)⋅ϕψ(v))⋅βα2(δ))⋅βα3(γ)(by(3.6))=(−1)|γ||δ|+|ε||v|((β2(ε)⋅ϕψ(v))⋅βα2(γ))⋅βα3(δ)+(−1)|ε||v|(αβ2(ε)⋅(ϕψ(v)⋅α2(δ)))⋅βα3(γ)(by(3.12))=(−1)|γ||δ|+|ε||v|((β2(ε)⋅ϕψ(v))⋅βα2(γ))⋅βα3(δ)⏟J1+(−1)|ε||v|α2β2(ε)⋅((ϕψ(v)⋅α2(δ))⋅α3(γ))⏟J2.(by(3.12)) |
B=(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|β(αβ(γ))α(αβ(δ))⋅ψ(ϕ2(v)⋅α3β−1(ε))=(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|α2β2(γ)⋅(α2β(δ)⋅(ϕ2(v)⋅α3β−1(ε)))+(−1)|γ||v|+|γ||ε|+|δ||v|+|δ||ε|α2β2(δ)⋅(α2β(γ)⋅(ϕ2(v)⋅α3β−1(ε)))(by(3.8))=(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|α2β2(γ)⋅((αβ(δ)ϕ2(v))⋅α3(ε))+(−1)|γ||v|+|γ||ε|+|δ||v|+|δ||ε|α2β2(δ)⋅(α2β(γ)⋅(ϕ2(v)⋅α3β−1(ε)))(by(3.12))=(−1)|γ||δ|+|γ||v|+|γ||ε|+|δ||v|+|δ||ε|(αβ2(γ)⋅(αβ(δ)⋅ϕ2(v)))⋅α3β(ε)⏟B1+(−1)|γ||v|+|γ||ε|+|δ||v|+|δ||ε|α2β2(δ)⋅(α2β(γ)⋅(ϕ2(v)⋅α3β−1(ε)))⏟B2.(by(3.12)) |
We substitute J1+J2 and B1+B2 for J and B to obtain
(−1)|γ||δ|asVϕ,ψ(ρr(ψ2(v),βα(ε)),βα2(γ),α3(δ))=(−1)|γ||δ|((ψ2(v)⋅βα(ε))⋅βα2(γ))⋅α3β(δ)+(−1)|δ||ε|+|δ||v|α2β2(δ)⋅((ϕψ(v)⋅α2(ε))⋅α3(γ))+(−1)|γ||δ|+|γ||v|+|ε||γ|(β2α(γ)⋅(ϕψ(v)⋅α2(ε)))⋅α3β(δ)+(−1)|ε||v|+|δ||ε|+|δ||v|α2β2(δ)⋅((αβ(ε)⋅ϕ2(v))α3(γ))+(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|(β2α(γ)⋅(αβ(ε)⋅ϕ2(v)))⋅α3β(δ)+(−1)|ε||v|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|α2β2(δ)⋅(α2β(γ)⋅(α2(ε)⋅ϕ3ψ−1(v)))−(−1)|γ||δ|(ϕψ2(v)⋅βα2(ε))⋅βα2(γ)α3(δ)−B1−J2−(−1)|γ||δ|+|ε||v|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|β2α(γ)βα2(δ)⋅(α2β(γ)⋅ϕ3(v)). |
Similarly, we have
(−1)|ε||γ|+|ε||δ|asVϕ,ψ(ρr(ψ2(v)⋅βα(δ)),βα2(γ),α3(ε))=(−1)|ε||γ|+|ε||δ|((ψ2(v)⋅βα(δ))⋅βα2(γ))⋅α3β(ε)+(−1)|ε||v|α2β2(ε)⋅((ϕψ(v)⋅α2(δ))⋅α3(γ))+(−1)|γ||v|+|δ||γ|+|ε||γ|+|ε||δ|(β2α(γ)⋅(ϕψ(v)⋅α2(δ)))⋅α3β(ε)+(−1)|δ||v|+|ε||v|α2β2(ε)⋅((αβ(δ)ϕ2(v))⋅α3(γ))+(−1)|δ||v|+|γ||δ|+|γ||v|+|ε||δ|+|ε||γ|(β2α(γ)⋅(αβ(δ)⋅ϕ2(v)))⋅α3β(ε)+(−1)|δ||v|+|γ||δ|+|γ||v|+|ε||v|α2β2(ε)⋅(α2β(γ)⋅(α2(δ)⋅ϕ3ψ−1(v)))−(−1)|ε||γ|+|ε||δ|(ϕψ2(v)⋅βα2(δ))⋅βα2(γ)α3(ε)−(−1)|ε||v|+|δ||v|+|ε||γ|+|ε||δ|α2β2(δ)⋅((ϕψ(v)⋅α2(ε))⋅α3(γ))−(−1)|γ||v|+|γ||δ|+|ε||v|+|ε||γ|(αβ2(γ)⋅(αβ(ε)⋅ϕ2(v)))⋅α3β(δ)−(−1)|ε||v|+|δ||v|+|γ||δ|+|γ||v|+|ε||γ|β2α(γ)α2β(ε)⋅(α2β(δ)⋅ϕ3(v)). |
In addition,
(−1)|v||ε|+|v||γ|+|v||δ|asVϕ,ψ(β2(ε)βα(δ),βα2(γ),ϕ3(v))=(−1)|v||ε|+|v||γ|+|v||δ|ρl((β2(ε)βα(δ))βα2(γ),ϕ3ψ(v))−(−1)|v||ε|+|v||γ|+|v||δ|ρl(β2α(ε)βα2(δ),ρl(βα2(γ),ϕ3(v))=(−1)|v||ε|+|v||γ|+|v||δ|((β2(ε)βα(δ))βα2(γ))⋅ϕ3ψ(v)+ϕ2ψ2(v)⋅((αβ(ε)α2(δ))α3(γ))−(−1)|v||ε|+|v||γ|+|v||δ|ρl(β2α(ε)βα2(δ),βα2(γ)⋅ϕ3(v))−(−1)|v||ε|+|v||δ|ρl(β2α(ε)βα2(δ),ψϕ2(v)⋅α3(γ))(by(3.13))=(−1)|v||ε|+|v||γ|+|v||δ|((β2(ε)βα(δ))βα2(γ))⋅ϕ3ψ(v)+ϕ2ψ2(v)⋅((αβ(ε)α2(δ))α3(γ))−(−1)|v||ε|+|v||γ|+|v||δ|(β2α(ε)βα2(δ))⋅(βα2(γ)⋅ϕ3(v))−(−1)|v||γ|+|γ||ε|+|γ||δ|(β2α(γ)ϕ2ψ(v))⋅βα2(ε)α3(δ)⏟D−(−1)|v||ε|+|v||δ|β2α(ε)βα2(δ)⋅(ψϕ2(v)⋅α3(γ))⏟C−(−1)|γ||ε|+|γ||δ|(ψ2ϕ(v)⋅α2β(γ))⋅α2β(ε)α3(δ).(by(3.13)) |
The same way, we replace C and D as follows
C=(−1)|v||ε|+|v||δ|β(αβ(ε))α(αβ(δ))⋅ψ(ϕ2(v)⋅α3β−1(γ))=(−1)|v||ε|+|v||δ|β2α2(ε)⋅(α2β(δ)⋅(ϕ2(v)⋅α3β−1(γ)))+(−1)|v||ε|+|v||δ|+|ε||δ|β2α2(δ)⋅(α2β(ε)⋅(ϕ2(v)⋅α3β−1(γ)))(by3.8)=(−1)|v||ε|+|v||δ|β2α2(ε)⋅((αβ(δ)⋅ϕ2(v))⋅α3(γ))⏟C1+(−1)|v||ε|+|v||δ|+|ε||δ|β2α2(δ)⋅((αβ(ε)⋅ϕ2(v))⋅α3(γ))⏟C2,(by3.12) |
D=(−1)|v||γ|+|γ||ε|+|γ||δ|ϕ(β2(γ)⋅ϕψ(v))⋅β(α2(ε))α(α2(δ))=(−1)|v||γ|+|γ||ε|+|γ||δ|((β2(γ)⋅ϕψ(v))⋅α2β(ε))⋅βα3(δ)+(−1)|v||γ|+|γ||ε|+|γ||δ|+|ε||δ|((β2(γ)⋅ϕψ(v))⋅α2β(δ))⋅βα3(ε)(by3.6)=(−1)|v||γ|+|γ||ε|+|γ||δ|(αβ2(γ)⋅(ϕψ(v)⋅α2(ε)))⋅βα3(δ)⏟D1+(−1)|v||γ|+|γ||ε|+|γ||δ|+|ε||δ|(αβ2(γ)⋅(ϕψ(v)⋅α2(δ)))⋅βα3(ε)⏟D2.(by3.12) |
Finally, we have
(−1)|γ||δ|asVϕ,ψ(ψ2(v)⋅αβ(ε),βα2(γ),α3(δ))+(−1)|ε||γ|+|ε||δ|asVϕ,ψ(ψ2(v)⋅αβ(δ),βα2(γ),α3(ε))+(−1)|v||ε|+|v||γ|+|v||δ|asVϕ,ψ(μ(β2(ε),αβ(δ)),βα2(γ),ϕ3(v))=(−1)|γ||δ|(3.2)+(−1)|v||ε|+|γ||ε|+|γ||v|+|δ||ε|+|δ||v|(3.4)=0. |
Hence, we prove that (V,ϕ,ψ,ρl,ρr) is a Bihom-Jordan J-super-bimodule.
Lemma 3.1. Let (J,μ,α,β) be a Bihom-associative superalgebra and (V,ϕ,ψ) be a Bihom-super-module.
1) If (V,ϕ,ψ) is a right Bihom-associative J-super-module with the structure map ρr, then (V,ϕ,ψ) is a right special Bihom-Jordan J⋄-super-module with the same structure map ρr.
2) If (V,ϕ,ψ) is a left Bihom-associative J-super-module with the structure map ρl such that ψ is invertible, then (V,ϕ,ψ) is a left special Bihom-Jordan J⋄-super-module with the same structure map ρl.
Proof. It also suffices to prove Eqs (3.6) and (3.8).
1) If (V,ϕ,ψ) is a right Bihom-associative J-super-module with the structure map ρr then for all (ε,γ,v)∈H(J)×H(J)×H(V). ϕ(v)⋅(β(ε)⋄α(γ))=ϕ(v)⋅(β(ε)α(γ)+(−1)|ε||γ|β(γ)α(ε))=(v⋅β(ε))⋅αβ(γ)+(−1)|ε||γ|(v⋅β(γ))⋅αβ(ε). Then (V,ϕ,ψ) is a right special Bihom-Jordan J⋄-super-module by Theorem 3.1.
2) Similarly, it is easy to obtain by Theorem 3.2.
End of lemma proof.
By Lemma 3.1 and Theorem 3.4, we obtain the following conclusion.
Proposition 3.1. Let (J,μ,α,β) be a Bihom-associative superalgebra and (V,ϕ,ψ,ρ1,ρ2) be a Bihom-associative J-super-bimodule such that ϕ and ψ are inversible. Then (V,ϕ,ψ,ρl,ρr) is a Bihom-Jordan J⋄-super-bimodule where ρl and ρr are defined as in Eq (3.13).
That is, a Bihom-associative J-super-bimodule gives rise to a Bihom-Jordan super-bimodule for J⋄.
Proposition 3.2. Let (J,μ,α,β) be a Bihom-Jordan superalgebra and Vϕ,ψ=(V,ϕ,ψ,ρl,ρr) be a Bihom-Jordan J-super-bimodule. Then for each n∈N such that ϕn=ψn=IdV, the maps
ρ(n)l=ρl∘(αn⊗ψn), | (3.14) |
and
ρ(n)r=ρr∘(ϕn⊗βn). | (3.15) |
as structure maps, (V,ϕ,ψ,ρ(n)l,ρ(n)r) is given to be a Bihom-Jordan J-super-bimodule. Denoted it by V(n)ϕ,ψ.
Proof. ρl and ρn are easy to prove special left and right super-modules, respectively, which are also left and right super-modules, and Eq (3.9) holds in V(n)ϕ,ψ. By direct calculation, we can convert asV(n)ϕ,ψ in V(n)ϕ,ψ to asVϕ,ψ in Vϕ,ψ, that is
asV(n)ϕ,ψ(β2(ε)αβ(γ),ϕ2ψ(v),α3(δ))=asVϕ,ψ(β2(αn(ε))αβ(αn(γ)),ϕ2ψ(v),α3(βn(δ))) |
, furthermore, we have
↺ε,γ,δ(−1)|δ|(|ε|+|v|)asV(n)ϕ,ψ(β2(ε)αβ(γ),ϕ2ψ(v),α3(δ))=↺αn(ε),αn(γ),βn(δ)(−1)|δ|(|ε|+|v|)asVϕ,ψ(β2(αn(ε))αβ(αn(γ)),ϕ2ψ(v),α3(βn(δ)))=0. |
Then we obtain Eq (3.10) in V(n)ϕ,ψ. Similarly, Eq (3.11) also holds in V(n)ϕ,ψ, which implies that V(n)ϕ,ψ is a Bihom-Jordan J-super-bimodule.
In the sequel, we present some results of Bihom-Jordan super-bimodules constructed by Jordan super-bimodules via endomorphisms.
Theorem 3.5. Let (J,μ) be a Jordan superalgebra, (V,ρl,ρr) be a Jordan J-super-bimodule, α,β be endomorphisms of J, which satisfies αβ=βα and ϕ,ψ be even linear self-maps of V such that ϕ∘ρl=ρl∘(α⊗ϕ), ϕ∘ρr=ρr∘(ϕ⊗α), ψ∘ρl=ρl∘(β⊗ψ) and ψ∘ρr=ρr∘(ψ⊗β). Denote Jα,β for the Bihom-Jordan superalgebra (J,μα,β=μ(α⊗β),α,β) and Vϕ,ψ for the Bihom-super-module (V,ϕ,ψ). Define two structure maps as follows:
~ρl=ρl(α⊗ψ)and~ρr=ρr(ϕ⊗β). | (3.16) |
Then Vϕ,ψ=(V,ϕ,ψ,~ρl,~ρr) is a Bihom-Jordan Jα,β-super-bimodule.
Proof. By direct calculation, it is easy to get asVϕ,ψ(μα,β(β2(ε),αβ(γ)),ϕ2ψ(v),α3(δ))=asV(α3β2(ε)α3β2(γ),ϕ3ψ2(v),α3β2(δ)), So it is clear Eqs (3.10) and (3.11) hold in Vϕ,ψ. Thus, (V,ϕ,ψ,~ρl,~ρr) is a Bihom-Jordan Jα,β-super-bimodule.
From Proposition 3.2 and Theorem 3.5, we have the following
Corollary 3.1. Let (J,μ) be a Jordan superalgebra, (V,ρl,ρr) be a Jordan J-super-bimodule, α,β be endomorphisms of J, which satify αβ=βα and ϕ,ψ be even linear self-maps of V such that ϕ∘ρl=ρl∘(α⊗ϕ), ϕ∘ρr=ρr∘(ϕ⊗α), ψ∘ρl=ρl∘(β⊗ψ) and ψ∘ρr=ρr∘(ψ⊗β). Denote Jα,β for the Bihom-Jordan superalgebra (J,μα,β=μ(α⊗β),α,β) and Vϕ,ψ for the Bihom-super-module (V,ϕ,ψ). Define two structure maps as follows:
~ρl(n)=ρl(αn+1⊗ψ)and~ρr(n)=ρr(ϕ⊗βn+1). | (3.17) |
Then Vϕ,ψ=(V,ϕ,ψ,~ρl(n),~ρr(n)) is a Bihom-Jordan Jα,β-super-bimodule for each n∈N.
Definition 3.3. An abelian extension of Bihom-Jordan superalgebra is a short exact sequence of Bihom-Jordan superalgebra:
0⟶(V,ϕ,ψ)i⟶(J,μJ,αJ,βJ)π⟶(B,μB,αB,βB)⟶0. |
where (V,ϕ,ψ) is a trivial Bihom-Jordan superalgebra, i and π are even morphisms of Bihom-superalgebras. If there exists an even morphism s:(B,μB,αB,βB)→(J,μJ,αJ,βJ) satisfies π∘s=IdB. Then the abelian extension is said to be split and s is called a section of π.
In this section, we study the representation and O-operator. Meanwhile, we characterize Bihom-pre-Jordan superalgebras by using O-operator.
Definition 4.1. Let (J,μ,α,β) be a Bihom-Jordan superalgebra, V be a Z2-graded vector spaces, ρ:J→End(V), ϕ,ψ∈Aug(V). Then (V,ρ,ϕ,ψ) is a representation of (J,μ,α,β), if the following conditions hold:
ϕψ=ψϕ, | (4.1) |
ρ(μ(μ(β2(ε),αβ(γ)),α2β(δ)))ϕ3ψ+(−1)|δ||γ|ρ(α2β2(ε))ϕψ−1ρ(αβ2(δ))ϕψ−1ρ(β2(γ))ϕψ+(−1)|γ||ε|+|δ||ε|ρ(α2β2(γ))ϕψ−1ρ(αβ2(δ))ϕψ−1ρ(β2(ε))ϕψ−ρ(μ(αβ2(ε),α2β(γ)))ρ(α2β(δ))ϕ3−(−1)|ε||δ|+|δ||γ|ρ(μ(αβ2(δ),α2β(ε)))ϕ2ψ−1ρ(β2(γ))ϕψ−(−1)|ε||δ|+|δ||γ|+|ε||γ|ρ(μ(αβ2(δ),α2β(γ)))ϕ2ψ−1ρ(β2(ε))ϕψ=0. | (4.2) |
↺ε,γ,δ(−1)|ε||δ|ρ(α2β2(ε))ϕψ−1ρ(μ(β2(γ),αβ(δ)))ϕ2ψ=↺ε,γ,δ(−1)|ε||δ|ρ(μ(αβ2(ε),α2β(γ)))ρ(α2β(δ))ϕ3. | (4.3) |
Example 4.1. Let (J,μ,α,β) be a regular Bihom-Jordan superalgebra. Define ad:J→End(J), for any ε,γ∈H(J), ad(ε)γ=μ(ε,γ). Then (J,ad,α,β) is a representation of (J,μ,α,β), which is called adjoint representation.
Proposition 4.1. Let (J,μ,α,β) be a Bihom-Jordan superalgebra. (V,ρ,ϕ,ψ) be a representation, define an even bilinear map μ and two even linear maps α and β on J⊕V as follows: for any ε,γ∈H(J),a,b∈H(V),
μ′(ε+a,γ+b)=μ(ε,γ)+ρ(ε)b+ρ(α−1β(γ))ϕψ−1(a), |
(α+ϕ)(ε+a)=α(ε)+ϕ(a),(β+ψ)(ε+a)=β(a)+ψ(a). |
Then (J⊕V,μ′,α+ϕ,β+ψ) is a Bihom-Jordan superalgebra, denoted by J⋉V and called semidirect product.
Proof. It can be verified directly by Definition 4.1.
We also consider the split null extension on J⋉V in Proposition 4.1.
Remark 4.1. Write elements a+v of J⊕V as (a,v). There is an injective homomorphism and a surjective homomorphism of Bihom-modules, respectively, as follows:
● i:V→J⊕V, i(v)=(0,v),
● π:J⊕V→, π(a,v)=a.
Moreover, i(V) is a Bihom-ideal of J⊕V such that J⊕V/i(V)≅J. On the other hand, there is an even morphisms σ:J→J⊕V given by σ(a)=(a,0), which is clearly a section of π. Therefore, we obtain the abelian split exact sequence:
![]() |
Definition 4.2. A BiHom superalgebra (J,⋅,α,β) is called a Bihom-pre-Jordan superalgebra if for all ε,γ,δ,t∈H(J):
1) αβ=βα, both α and β are reversible,
1)
((β2(ε)⋅αβ(γ))⋅α2β(δ))⋅α3β(w)+(−1)|ε||γ|((β2(γ)⋅αβ(ε))⋅α2β(δ))⋅α3β(w)+(−1)|δ|(|ε|+|γ|)(αβ2(δ)⋅(αβ(ε)α2(γ)))⋅α3β(w)+(−1)|δ|(|ε|+|γ|)+|ε||γ|(αβ2(δ)⋅(αβ(γ)α2(ε)))⋅α3β(w)+(−1)|δ||γ|α2β2(ε)⋅(α2β(δ)⋅(α2(γ)⋅α3β−1(w)))+(−1)|δ||ε|+|γ||ε|α2β2(γ)⋅(α2β(δ)⋅(α2(ε)⋅α3β−1(w)))−(αβ2(ε)⋅α2β(γ))⋅(α2β(δ)⋅α3(w))−(−1)|ε||γ|(αβ2(γ)⋅α2β(ε))⋅(α2β(δ)⋅α3(w))−(−1)|ε||δ|+|δ||γ|(αβ2(δ)⋅α2β(ε))⋅(α2β(γ)⋅α3(w))−(−1)|δ||γ|(αβ2(ε)⋅α2β(δ))⋅(α2β(γ)⋅α3(w))−(−1)|γ||δ|+|δ||ε|+|γ||ε|(αβ2(δ)⋅α2β(γ))⋅(α2β(ε)⋅α3(w))−(−1)|δ||ε|+|γ||ε|(αβ2(γ)⋅α2β(δ))⋅(α2β(ε)⋅α3(w))=0, | (4.4) |
3)
(−1)|w||γ|α2β2(w)⋅((αβ(ε)⋅α2(γ))⋅α3(δ))+(−1)|ε||γ|+|w||γ|α2β2(w)⋅((αβ(γ)⋅α2(ε))⋅α3(δ))+(−1)|ε||w|α2β2(ε)⋅((αβ(γ)⋅α2(w))⋅α3(δ))+(−1)|ε||w|+|γ||w|α2β2(ε)⋅((αβ(w)⋅α2(γ))⋅α3(δ))+(−1)|γ||ε|α2β2(γ)⋅((αβ(w)⋅α2(ε))⋅α3(δ))+(−1)|γ||ε|+|w||ε|α2β2(γ)⋅((αβ(ε)⋅α2(w))⋅α3(δ))−(−1)|w||ε|(αβ2(ε)⋅α2β(γ))⋅(α2β(w)⋅α3(δ))−(−1)|ε||γ|+|w||ε|(αβ2(γ)⋅α2β(ε))⋅(α2β(w)⋅α3(δ))−(−1)|ε||γ|(αβ2(γ)⋅α2β(w))⋅(α2β(ε)⋅α3(δ))−(−1)|γ||w|+|ε||γ|(αβ2(w)⋅α2β(γ))⋅(α2β(ε)⋅α3(δ))−(−1)|γ||w|(αβ2(w)⋅α2β(ε))⋅(α2β(γ)⋅α3(δ))−(−1)|w||ε|+|γ||w|(αβ2(ε)⋅α2β(w))⋅(α2β(γ)⋅α3(δ)). | (4.5) |
Actually, condition 3 is equivalent to
↺ε,γ,w{(−1)|ε||w|α2β2(ε)⋅((αβ(γ)⋅α2(w))⋅α3(δ))+(−1)|ε||w|+|γ||w|α2β2(ε)⋅((αβ(w)⋅α2(γ))⋅α3(δ))}=↺ε,γ,w{(−1)|ε||w|(αβ2(ε)⋅α2β(γ))⋅(α2β(w)⋅(δ))+(−1)|ε||w|+|ε||γ|(αβ2(γ)⋅α2β(ε))⋅(α2β(w)⋅(δ))}. |
Theorem 4.1. Let (J,⋅,α,β) be a Bihom-pre-Jordan superalgebra, define an even bilinear operator μ: for all ε,γ∈H(J)
μ(ε,γ)=ε⋅γ+(−1)|ε||γ|α−1β(γ)⋅αβ−1(ε), | (4.6) |
then (J,∙,α,β) is a Bihom-Jordan superalgebra.
Proof. By Eq (4.6), we get
μ(β(ε),α(γ))=β(ε)⋅α(γ)+(−1)|ε||γ|β(γ)⋅α(ε)=(−1)|ε||γ|μ(β(γ),α(ε)). |
That is to say the Bihom-super commutativity condition holds. Next, by direct calculation,
(−1)|w|(|ε|+|δ|)~asα,β(μ(β2(ε),αβ(γ)),α2β(δ),α3(w))=(−1)|w|(|ε|+|δ|)((β2(ε)⋅αβ(γ))⋅α2β(δ))⋅α3β(w)_①+(−1)|w||γ|α2β2(w)⋅((αβ(ε)⋅α2(γ))⋅α3(δ))⏟1′+(−1)|w|(|ε|+|δ|)+|δ|(|ε|+|γ|)(αβ2(δ)⋅(αβ(ε)⋅α2(γ)))⋅α3β(w)_②+(−1)|w||γ|+|δ|(|ε|+|γ|)α2β2(w)⋅(α2β(δ)⋅(α2(ε)⋅α3β−1(γ)))−(−1)|w|(|ε|+|δ|)(αβ2(ε)⋅α2β(γ))⋅(α2β(δ)⋅α3(w))_③−(−1)|w|(|δ|+|γ|)+|δ|(|ε|+|γ|)(αβ2(δ)⋅α2β(w))⋅(α2β(ε)⋅α3(γ))−(−1)|w||ε|(αβ2(ε)⋅α2β(γ))⋅(α2β(w)⋅α3(δ))⏟2′−(−1)|w||γ|+|δ|(|ε|+|γ|)(αβ2(w)⋅α2β(δ))⋅(α2β(ε)⋅α3(γ))+(−1)|w|(|ε|+|δ|)+|ε||γ|((β2(γ)⋅αβ(ε))⋅α2β(δ))⋅α3β(w)_④+(−1)|w||γ|+|ε||γ|α2β2(w)⋅((αβ(γ)⋅α2(ε))⋅α3(δ))⏟3′+(−1)|w|(|ε|+|δ|)+|δ|(|ε|+|γ|)+|ε||γ|(αβ2(δ)⋅(αβ(γ)⋅α2(ε)))⋅α3β(w)_⑤+(−1)|w||γ|+|δ|(|ε|+|γ|)+|ε||γ|α2β2(w)⋅(α2β(δ)⋅(α2(γ)⋅α3β−1(ε)))−(−1)|w|(|ε|+|δ|)+|ε||γ|(αβ2(γ)⋅α2β(ε))⋅(α2β(δ)⋅α3(w))_⑥−(−1)|w|(|δ|+|γ|)+|δ|(|ε|+|γ|)+|ε||γ|(αβ2(δ)⋅α2β(w))⋅(α2β(γ)⋅α3(ε))−(−1)|w||ε|+|ε||γ|(αβ2(γ)⋅α2β(ε))⋅(α2β(w)⋅α3(δ))⏟4′−(−1)|w||γ|+|δ|(|ε|+|γ|)+|ε||γ|(αβ2(w)⋅α2β(δ))⋅(α2β(γ)⋅α3(ε)), |
(−1)|ε|(|γ|+|δ|)~asα,β(μ(β2(γ),αβ(w)),α2β(δ),α3(ε))=(−1)|ε|(|γ|+|δ|)((β2(γ)⋅αβ(w))⋅α2β(δ))⋅α3β(ε)+(−1)|ε||w|α2β2(ε)⋅((αβ(γ)⋅α2(w))⋅α3(δ))⏟5′+(−1)|ε|(|γ|+|δ|)+|δ|(|w|+|γ|)(αβ2(δ)⋅(αβ(γ)⋅α2(w)))⋅α3β(ε)+(−1)|ε||w|+|δ|(|γ|+|w|)α2β2(ε)⋅(α2β(δ)⋅(α2(γ)⋅α3β−1(w)))_⑦−(−1)|ε|(|γ|+|δ|)(αβ2(γ)⋅α2β(w))⋅(α2β(δ)⋅α3(ε))−(−1)|ε|(|δ|+|w|)+|δ|(|γ|+|w|)(αβ2(δ)⋅α2β(ε))⋅(α2β(γ)⋅α3(w))_⑧−(−1)|ε||γ|(αβ2(γ)⋅α2β(w))⋅(α2β(ε)⋅α3(δ))⏟6′−(−1)|ε||w|+|δ|(|γ|+|w|)(αβ2(ε)⋅α2β(δ))⋅(α2β(γ)⋅α3(w))_⑨+(−1)|ε|(|γ|+|δ|)+|γ||w|((β2(w)⋅αβ(γ))⋅α2β(δ))⋅α3β(ε)+(−1)|ε||w|+|γ||w|α2β2(ε)⋅((αβ(w)⋅α2(γ))⋅α3(δ))⏟7′+(−1)|ε|(|γ|+|δ|)+|δ|(|γ|+|w|)+|γ||w|(αβ2(δ)⋅(αβ(w)⋅α2(γ)))⋅α3β(ε)+(−1)|w||γ|+|δ|(|γ|+|w|)+|ε||w|α2β2(ε)⋅(α2β(δ)⋅(α2(w)⋅α3β−1(γ)))−(−1)|ε|(|γ|+|δ|)+|γ||w|(αβ2(w)⋅α2β(γ))⋅(α2β(δ)⋅α3(ε))−(−1)|ε|(|δ|+|w|)+|δ|(|γ|+|w|)+|γ||w|(αβ2(δ)⋅α2β(ε))⋅(α2β(w)⋅α3(γ))−(−1)|w||γ|+|ε||γ|(αβ2(w)⋅α2β(γ))⋅(α2β(ε)⋅α3(δ))⏟8′−(−1)|w||γ|+|δ|(|γ|+|w|)+|ε||w|(αβ2(ε)⋅α2β(δ))⋅(α2β(w)⋅α3(γ)), |
(−1)|γ|(|w|+|δ|)~asα,β(μ(β2(w),αβ(ε)),α2β(δ),α3(γ))=(−1)|γ|(|w|+|δ|)((β2(w)⋅αβ(ε))⋅α2β(δ))⋅α3β(γ)+(−1)|γ||ε|α2β2(γ)⋅((αβ(w)⋅α2(ε))⋅α3(δ))⏟9′+(−1)|γ|(|w|+|δ|)+|δ|(|w|+|ε|)(αβ2(δ)⋅(αβ(w)⋅α2(ε)))⋅α3β(γ)+(−1)|γ||ε|+|δ|(|w|+|ε|)α2β2(γ)⋅(α2β(δ)⋅(α2(w)⋅α3β−1(ε)))−(−1)|γ|(|w|+|δ|)(αβ2(w)⋅α2β(ε))⋅(α2β(δ)⋅α3(γ))−(−1)|γ|(|δ|+|ε|)+|δ|(|w|+|ε|)(αβ2(δ)⋅α2β(γ))⋅(α2β(w)⋅α3(ε))−(−1)|γ||w|(αβ2(w)⋅α2β(ε))⋅(α2β(γ)⋅α3(δ))⏟10′−(−1)|γ||ε|+|δ|(|w|+|ε|)(αβ2(γ)⋅α2β(δ))⋅(α2β(w)⋅α3(ε))+(−1)|γ|(|w|+|δ|)+|w||ε|((β2(ε)⋅αβ(w))⋅α2β(δ))⋅α3β(γ)+(−1)|w||ε|+|ε||γ|α2β2(γ)⋅((αβ(ε)⋅α2(w))⋅α3(δ))⏟11′+(−1)|γ|(|w|+|δ|)+|δ|(|w|+|ε|)+|w||ε|(αβ2(δ)⋅(αβ(ε)⋅α2(w)))⋅α3β(γ)+(−1)|w||ε|+|δ|(|w|+|ε|)+|ε||γ|α2β2(γ)⋅(α2β(δ)⋅(α2(ε)⋅α3β−1(w)))_⑩−(−1)|γ|(|w|+|δ|)+|w||ε|(αβ2(ε)⋅α2β(w))⋅(α2β(δ)⋅α3(γ))−(−1)|γ|(|δ|+|ε|)+|δ|(|w|+|ε|)+|w||ε|(αβ2(δ)⋅α2β(γ))⋅(α2β(ε)⋅α3(w))_⑪−(−1)|w||ε|+|γ||w|(αβ2(ε)⋅α2β(w))⋅(α2β(γ)⋅α3(δ))⏟12′−(−1)|w||ε|+|δ|(|w|+|ε|)+|ε||γ|(αβ2(γ)⋅α2β(δ))⋅(α2β(ε)⋅α3(w))_⑫, |
By Eq (4.4), we have ①+⋯+⑫=0, and by Eq (4.5), 1′+⋯+12′=0, Analogously, the conclusion that the sum is zero can be obtained by recombining the remaining unmarked formulas, which implies
↺ε,γ,w(−1)|w|(|ε|+|δ|)~asα,β(μ(β2(ε),αβ(γ)),α2β(δ),α3(w))=0. |
This completes the proof.
Definition 4.3. Let (J,μ,α,β) be a Bihom-Jordan superalgebra, and (V,ρ,ϕ,ψ) be its representation. If even the linear map T:J→V satisfies the following conditions: for all a,b∈H(V),
μ(T(a),T(b))=T(ρ(T(a))b+(−1)|a||b|ρ(T(ϕ−1ψ(b)))ϕψ−1(a)), |
T∘ϕ=α∘T,T∘ψ=β∘T, |
then T is called O-operator with respect to representation.
Definition 4.4. Let (J,μ,α,β) be a Bihom-Jordan superalgebra and α,β be reversible, R∈gl(J), R is called Rota–Baxter operator on J, if for all ε,γ∈H(J), the following conditions hold:
μ(R(ε),R(γ))=R(μ(R(ε),γ)+(−1)|ε||γ|μ(R(α−1β(γ)),αβ−1(ε))), |
R∘α=α∘R,R∘β=β∘R. |
Theorem 4.2. Let (J,μ,α,β) be a Bihom-Jordan superalgebra, (V,ρ,ϕ,ψ) be its representation, and T be an O-operator with respect to representation. Define bilinear operation ⋅ on V:
a⋅b=ρ(T(a))b,∀a,b∈H(V). |
Then (V,⋅,ϕ,ψ) is a Bihom-pre-Jordan superalgebra.
Proof. Actually, it can be calculated directly from Definition 4.1.
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
This research is supported by NNSF of China (Nos. 12271085 and 12071405). The authors would like to thank the reviewers for valuable suggestions to improve the paper.
The authors declare there are no conflicts of interest.
[1] |
Y. C. Wang, C. H. Cheng, A multiple combined method for rebalancing medical data with class imbalances, Comput. Biol. Med., 134 (2021), 104527. https://doi.org/10.1016/j.compbiomed.2021.104527 doi: 10.1016/j.compbiomed.2021.104527
![]() |
[2] |
M. Zareapoor, P. Shamsolmoali, J. Yang, Oversampling adversarial network for class-imbalanced fault diagnosis, Mech. Syst. Signal Process., 149 (2021), 107175. https://doi.org/10.1016/j.ymssp.2020.107175 doi: 10.1016/j.ymssp.2020.107175
![]() |
[3] |
S. Fan, X. Zhang, Z. Song, Imbalanced sample selection with deep reinforcement learning for fault diagnosis, IEEE Trans. Ind. Inf., 18 (2021), 2518–2527. https://doi.org/10.1109/TⅡ.2021.3100284 doi: 10.1109/TⅡ.2021.3100284
![]() |
[4] |
N. Gupta, V. Jindal, P. Bedi, LIO-IDS: Handling class imbalance using LSTM and improved one-vs-one technique in intrusion detection system, Comput. Networks, 192 (2021), 108076. https://doi.org/10.1016/j.comnet.2021.108076 doi: 10.1016/j.comnet.2021.108076
![]() |
[5] |
Z. Li, M. Huang, G. Liu, C. Jiang, A hybrid method with dynamic weighted entropy for handling the problem of class imbalance with overlap in credit card fraud detection, Expert Syst. Appl., 175 (2021), 114750. https://doi.org/10.1016/j.eswa.2021.114750 doi: 10.1016/j.eswa.2021.114750
![]() |
[6] |
A. G. C. de Sá, A. C. M. Pereira, G. L. Pappa, A customized classification algorithm for credit card fraud detection, Eng. Appl. Artif. Intell., 72 (2018), 21–29. https://doi.org/10.1016/j.engappai.2018.03.011 doi: 10.1016/j.engappai.2018.03.011
![]() |
[7] |
B. Guo, C. Zhang, J. Liu, X. Ma, Improving text classification with weighted word embeddings via a multi-channel TextCNN model, Neurocomputing, 363 (2019), 366–374. https://doi.org/10.1016/j.neucom.2019.07.052 doi: 10.1016/j.neucom.2019.07.052
![]() |
[8] |
Y. Li, H. Guo, Q. Zhang, M. Gu, J. Yang, Imbalanced text sentiment classification using universal and domain-specific knowledge, Knowl. Based Syst., 160 (2018), 1–15. https://doi.org/10.1016/j.knosys.2018.06.019 doi: 10.1016/j.knosys.2018.06.019
![]() |
[9] |
L. Dou, F. Yang, L. Xu, Q. Zou, A comprehensive review of the imbalance classification of protein post-translational modifications, Briefings Bioinf., 22 (2021), bbab089. https://doi.org/10.1093/bib/bbab089 doi: 10.1093/bib/bbab089
![]() |
[10] |
M. Neyestani, F. Sarmadian, A. Jafari, A. Keshavarzi, A. Sharififar, Digital mapping of soil classes using spatial extrapolation with imbalanced data, Geoderma Reg., 26 (2021), e00422. https://doi.org/10.1016/j.geodrs.2021.e00422 doi: 10.1016/j.geodrs.2021.e00422
![]() |
[11] |
S. Ketu, P. K. Mishra, Scalable kernel-based SVM classification algorithm on imbalance air quality data for proficient healthcare, Complex Intell. Syst., 7 (2021), 2597–2615. https://doi.org/10.1007/s40747-021-00435-5 doi: 10.1007/s40747-021-00435-5
![]() |
[12] |
Y. S. Li, H. Chi, X. Y. Shao, M. L. Qi, B. G. Xu, A novel random forest approach for imbalance problem in crime linkage, Knowl. Based Syst., 195 (2020), 105738. https://doi.org/10.1016/j.knosys.2020.105738 doi: 10.1016/j.knosys.2020.105738
![]() |
[13] |
P. Soltanzadeh, M. Hashemzadeh, RCSMOTE: Range-Controlled synthetic minority over-sampling technique for handling the class imbalance problem, Inf. Sci, 542 (2021), 92–111. https://doi.org/10.1016/j.ins.2020.07.014 doi: 10.1016/j.ins.2020.07.014
![]() |
[14] |
S. Susan, A. Kumar, The balancing trick: optimized sampling of imbalanced datasets—a brief survey of the recent State of the Art, Eng. Rep., 3 (2021), e12298. https://doi.org/10.1002/eng2.12298 doi: 10.1002/eng2.12298
![]() |
[15] |
H. Guan, Y. Zhang, M. Xian, H. D. Cheng, X. Tang, SMOTE-WENN: Solving class imbalance and small sample problems by oversampling and distance scaling, Appl. Intell., 51 (2021), 1394–1409. https://doi.org/10.1007/s10489-020-01852-8 doi: 10.1007/s10489-020-01852-8
![]() |
[16] |
H. Yu, C. Sun, X. Yang, S. Zheng, H. Zou, Fuzzy support vector machine with relative density information for classifying imbalanced data, IEEE Trans. Fuzzy Syst., 27 (2019), 2353–2367. https://doi.org/10.1109/TFUZZ.2019.2898371 doi: 10.1109/TFUZZ.2019.2898371
![]() |
[17] |
H. Zhang, L. Jiang, C. Li, CS-ResNet: Cost-sensitive residual convolutional neural network for PCB cosmetic defect detection, Expert Syst. Appl., 185 (2021), 115673. https://doi.org/10.1016/j.eswa.2021.115673 doi: 10.1016/j.eswa.2021.115673
![]() |
[18] |
W. Pei, B. Xue, L. Shang, M. Zhang, Genetic programming for development of cost-sensitive classifiers for binary high-dimensional unbalanced classification, Appl. Soft Comput., 101 (2021), 106989. https://doi.org/10.1016/j.asoc.2020.106989 doi: 10.1016/j.asoc.2020.106989
![]() |
[19] |
Z. H. Zhou, X. Y. Liu, Training cost-sensitive neural networks with methods addressing the class imbalance problem, IEEE Trans. Knowl. Data Eng., 18 (2006), 63–77. https://doi.org/10.1109/TKDE.2006.17 doi: 10.1109/TKDE.2006.17
![]() |
[20] |
W. J. Lin, J. J. Chen, Class-imbalanced classifiers for high-dimensional data, Briefings Bioinf., 14 (2012), 13–26. https://doi.org/10.1093/bib/bbs006 doi: 10.1093/bib/bbs006
![]() |
[21] |
H. Yu, C. Mu, C. Sun, W. Yang, X. Yang, X. Zuo, Support vector machine-based optimized decision threshold adjustment strategy for classifying imbalanced data, Knowl. Based Syst., 76 (2015), 67–78. https://doi.org/10.1016/j.knosys.2014.12.007 doi: 10.1016/j.knosys.2014.12.007
![]() |
[22] |
H. Yu, C. Sun, X. Yang, W. Yang, J. Shen, Y. Qi, ODOC-ELM: Optimal decision outputs compensation-based extreme learning machine for classifying imbalanced data, Knowl. Based Syst., 92 (2016), 55–70. https://doi.org/10.1016/j.knosys.2015.10.012 doi: 10.1016/j.knosys.2015.10.012
![]() |
[23] |
K. Yang, Z. Yu, C. L. P. Chen, W. Cao, J. You, H. S. Wong, Incremental weighted ensemble broad learning system for imbalanced data, IEEE Trans. Knowl. Data Eng., 34 (2021), 5809–5824. https://doi.org/10.1109/TKDE.2021.3061428 doi: 10.1109/TKDE.2021.3061428
![]() |
[24] |
Z. Qi, Z. Zhang, A hybrid cost-sensitive ensemble for heart disease prediction, BMC Med. Inf. Decis. Making, 21 (2021), 1–18. https://doi.org/10.21203/rs.2.22946/v1 doi: 10.21203/rs.2.22946/v1
![]() |
[25] |
H. Du, Y. Zhang, K. Gang, L. Zhang, Y. C. Chen, Online ensemble learning algorithm for imbalanced data stream, Appl. Soft Comput., 107 (2021), 107378. https://doi.org/10.1016/j.asoc.2021.107378 doi: 10.1016/j.asoc.2021.107378
![]() |
[26] |
T. Hayashi, H. Fujita, One-class ensemble classifier for data imbalance problems, Appl. Intell., 52 (2022), 17073–17089. https://doi.org/10.1007/s10489-021-02671-1 doi: 10.1007/s10489-021-02671-1
![]() |
[27] |
E. Schubert, J. Sander, M. Ester, H. P. Kriegel, X. Xu, DBSCAN revisited, revisited: why and how you should (still) use DBSCAN, ACM Trans. Database Syst., 42 (2017), 1–21. https://doi.org/10.1145/3068335 doi: 10.1145/3068335
![]() |
[28] |
T. N. Tran, K. Drab, M. Daszykowski, Revised DBSCAN algorithm to cluster data with dense adjacent clusters, Chemom. Intell. Lab. Syst., 120 (2013), 92–96. https://doi.org/10.1016/j.chemolab.2012.11.006 doi: 10.1016/j.chemolab.2012.11.006
![]() |
[29] |
D. Birant, A. Kut, ST-DBSCAN: An algorithm for clustering spatial–temporal data, Data Knowl. Eng., 60 (2007), 208–221. https://doi.org/10.1016/j.datak.2006.01.013 doi: 10.1016/j.datak.2006.01.013
![]() |
[30] |
M. Ontivero-Ortega, A. Lage-Castellanos, G. Valente, R. Goebel, M. Valdes-Sosa, Fast Gaussian Naïve Bayes for searchlight classification analysis, Neuroimage, 163 (2017), 471–479. https://doi.org/10.1016/j.neuroimage.2017.09.001 doi: 10.1016/j.neuroimage.2017.09.001
![]() |
[31] |
R. D. Raizada, Y. S. Lee, Smoothness without smoothing: why Gaussian Naive Bayes is not naive for multi-subject searchlight studies, PloS One, 8 (2013), e69566. https://doi.org/10.1371/journal.pone.0069566 doi: 10.1371/journal.pone.0069566
![]() |
[32] |
W. S. Noble, What is a support vector machine? Nat. Biotechnol. , 24 (2006), 1565–1567. https://doi.org/10.1038/nbt1206-1565 doi: 10.1038/nbt1206-1565
![]() |
[33] |
C. C. Chang, C. J. Lin, LIBSVM: A library for support vector machines, ACM Trans. Intell. Syst. Technol., 2 (2011), 1–27. https://doi.org/10.1145/1961189.1961199 doi: 10.1145/1961189.1961199
![]() |
[34] |
G. B. Huang, Q. Y. Zhu, C. K. Siew, Extreme learning machine: theory and applications, Neurocomputing, 70 (2006), 489–501. https://doi.org/10.1016/j.neucom.2005.12.126 doi: 10.1016/j.neucom.2005.12.126
![]() |
[35] |
G. B. Huang, H. Zhou, X. Ding, R. Zhang, Extreme learning machine for regression and multiclass classification, IEEE Trans. Syst. Man Cybern. Part B Cybern., 42 (2011), 513–529. https://doi.org/10.1109/tsmcb.2011.2168604 doi: 10.1109/tsmcb.2011.2168604
![]() |
[36] |
H. Yu, J. Ni, S. Xu, B. Qin, H. Jv, Estimating harmfulness of class imbalance by scatter matrix based class separability measure, Intell. Data Anal., 18 (2014), 203–216. https://doi.org/10.3233/IDA-140637 doi: 10.3233/IDA-140637
![]() |
[37] | P. E. Gill, W. Murray, M. H. Wright, Practical Optimization, Academic Press, London, 1981. https://doi.org/10.2307/3616583 |
[38] |
H. Guo, H. Liu, C. Wu, W. Zhi, Y. Xiao, W. She, Logistic discrimination based on G-mean and F-measure for imbalanced problem, J. Intell. Fuzzy Syst., 31 (2016), 1155–1166. https://doi.org/10.3233/ifs-162150 doi: 10.3233/ifs-162150
![]() |
[39] |
I. Triguero, S. González, J. M. Moyano, S. G. López, J. A. Fernández, J. L. Martín, KEEL 3.0: an open source software for multi-stage analysis in data mining, Int. J. Comput. Intell. Syst., 10 (2017), 1238–1249. https://doi.org/10.2991/ijcis.10.1.82 doi: 10.2991/ijcis.10.1.82
![]() |
[40] | J. Demšar, Statistical comparisons of classifiers over multiple data sets, J. Mach. Learn. Res., 7 (2006), 1–30. Available from: https://www.jmlr.org/papers/volume7/demsar06a/demsar06a.pdf. |
1. | Eric Avila-Vales, Ángel G. C. Pérez, Global properties of an age-structured virus model with saturated antibody-immune response, multi-target cells, and general incidence rate, 2021, 27, 1405-213X, 10.1007/s40590-021-00315-5 | |
2. | Yu Yang, Lan Zou, Yasuhiro Takeuchi, Global analysis of a multi-group viral infection model with age structure, 2020, 0003-6811, 1, 10.1080/00036811.2020.1721471 | |
3. | Huaqiao Zhang, Hong Chen, Cuicui Jiang, Kaifa Wang, Effect of explicit dynamics of free virus and intracellular delay, 2017, 104, 09600779, 827, 10.1016/j.chaos.2017.09.038 | |
4. | Shaoli Wang, Tengfei Wang, Yuming Chen, Bifurcations and Bistability of an Age-Structured Viral Infection Model with a Nonmonotonic Immune Response, 2022, 32, 0218-1274, 10.1142/S0218127422501516 | |
5. | Peng Wu, Hongyong Zhao, Mathematical analysis of multi-target cells and multi-strain age-structured model with two HIV infection routes, 2021, 14, 1793-5245, 10.1142/S1793524521500571 | |
6. | Junmei Liu, Yonggang Ma, Global stability of a HIV-1 CCR5 gene therapy with suicide gene, 2024, 12, 2164-2583, 10.1080/21642583.2023.2291407 | |
7. | A. Yu. Shcheglov, S. V. Netessov, An Inverse Problem for an Age-Structured Population Dynamics Model with Migration Flows, 2024, 17, 1995-4239, 93, 10.1134/S1995423924010099 | |
8. | S. V. Netesov, A. Yu. Shcheglov, Inverse Problem for a Nonlinear Model of Population Dynamics with the Age Structure of Individuals and Overpopulation, 2024, 48, 0278-6419, 20, 10.3103/S0278641924010072 |