Research article Special Issues

Linear stability analysis of overdetermined problems with non-constant data

  • Received: 17 March 2022 Revised: 16 July 2022 Accepted: 19 July 2022 Published: 09 August 2022
  • We study an overdetermined problem that arises as the Euler-Lagrange equation of a weighted variational problem in elasticity. Based on a detailed linear analysis by spherical harmonics, we prove the existence and local uniqueness as well as an optimal stability estimate for the shape of a domain allowing the solvability of the overdetermined problem. Our linear analysis reveals that the solution structure is strongly related to the choice of parameters in the problem. In particular, the global uniqueness holds for the pair of the parameters lying in a triangular region.

    Citation: Michiaki Onodera. Linear stability analysis of overdetermined problems with non-constant data[J]. Mathematics in Engineering, 2023, 5(3): 1-18. doi: 10.3934/mine.2023048

    Related Papers:

    [1] Bader S. Almohaimeed . Correction: Periodic stationarity conditions for mixture periodic INGARCH models. AIMS Mathematics, 2022, 7(10): 18280-18281. doi: 10.3934/math.20221005
    [2] Omar Alzeley, Ahmed Ghezal . On an asymmetric multivariate stochastic difference volatility: structure and estimation. AIMS Mathematics, 2024, 9(7): 18528-18552. doi: 10.3934/math.2024902
    [3] Kee Wah Fo, Seng Huat Ong, Choung Min Ng, You Beng Koh . An alternative hyper-Poisson integer-valued GARCH model with application to polio, internet protocol and COVID-19 data. AIMS Mathematics, 2023, 8(12): 29116-29139. doi: 10.3934/math.20231491
    [4] Chun Peng, Xiaoliang Li, Bo Du . Positive periodic solution for enterprise cluster model with feedback controls and time-varying delays on time scales. AIMS Mathematics, 2024, 9(3): 6321-6335. doi: 10.3934/math.2024308
    [5] Shuo Zhang, Jianhua Cheng . Bayesian premium of a credibility model based on a heterogeneous SETINAR(2, 1) process. AIMS Mathematics, 2023, 8(12): 28710-28727. doi: 10.3934/math.20231469
    [6] Lihong Guan, Xiaohong Wang . A discrete-time dual risk model with dependence based on a Poisson INAR(1) process. AIMS Mathematics, 2022, 7(12): 20823-20837. doi: 10.3934/math.20221141
    [7] Ahmed Ghezal, Mohamed balegh, Imane Zemmouri . Markov-switching threshold stochastic volatility models with regime changes. AIMS Mathematics, 2024, 9(2): 3895-3910. doi: 10.3934/math.2024192
    [8] Mashael A. Alshehri, Mohamed Kayid . Copulas generated by mixtures of weighted distributions. AIMS Mathematics, 2022, 7(5): 8953-8974. doi: 10.3934/math.2022499
    [9] Mohamed El-Borhamy, Essam Eddin M. Rashad, Arafa A. Nasef, Ismail Sobhy, Samah M. Elkholy . On the construction of stable periodic solutions for the dynamical motion of AC machines. AIMS Mathematics, 2023, 8(4): 8902-8927. doi: 10.3934/math.2023446
    [10] Tahani A. Abushal, Alaa H. Abdel-Hamid . Inference on a new distribution under progressive-stress accelerated life tests and progressive type-II censoring based on a series-parallel system. AIMS Mathematics, 2022, 7(1): 425-454. doi: 10.3934/math.2022028
  • We study an overdetermined problem that arises as the Euler-Lagrange equation of a weighted variational problem in elasticity. Based on a detailed linear analysis by spherical harmonics, we prove the existence and local uniqueness as well as an optimal stability estimate for the shape of a domain allowing the solvability of the overdetermined problem. Our linear analysis reveals that the solution structure is strongly related to the choice of parameters in the problem. In particular, the global uniqueness holds for the pair of the parameters lying in a triangular region.



    Integer-valued GARCH (INGARCH) models have proved to be useful in modeling count time series which are characterized by specific patterns that cannot be accounted for by standard ARMA models ([1,5,18,19,20,23,25,30,32,37,38,39]). Among the most apparent features of count time series are low values, overdispersion (the variance is larger than the mean), persistence, asymmetry, and a positive autocorrelation structure (cf. [19] and [27]). In its general form, the INGARCH model is defined through a discrete conditional distribution (e.g., Poisson, negative binomial, etc.) with a stochastic time-varying conditional mean, which depends on past observations through time-invariant parameters.

    For count data that also exhibit a seasonal behavior, the periodic INGARCH model – in which the parameters are taken to be periodic over time – has been found to be attractive ([7,14,16,29]). Aknouche et al. [7] proposed periodic stationarity and periodic ergodicity conditions for the first-order periodic INGARCH(1,1) model with conditional distributions belonging to the class of Poisson mixtures. Almohaimeed in [11] extended the result to a higher-order periodic INGARCH(p,q) in a larger family of distributions whose stochastic order is identical to the mean order ([3]). These general results can be used to support asymptotic inference for periodic INGARCH models used in studying the asymptotic properties of the Poisson quasi-maximum likelihood estimate (QMLE) and the negative binomial QMLE ([1] and [8]).

    In spite of the high-level of generality of the periodic INGARCH model in modeling a wide range of count data, it seems that it is unable to model some other pathological features such as multimodality and heavy tailedness of the conditional distributions ([21,22,40]). A well-known approach to model these features is to utilize a finite mixture of distributions (e.g., [4,10]) leading to a mixture periodic INGARCH representation. Ouzzani and Bentarzi [35] proposed a Poisson mixture periodic INARCH model and studied its periodic stationarity in mean properties. However, strict periodic stationarity and periodic ergodicity have not been considered. Moreover, the model only deals with the Poisson mixture and the ARCH forms, which can be restrictive.

    This paper provides strict periodic stationarity and periodic ergodicity conditions for a general mixture periodic INGARCH(p,q) model. This model is defined via a finite mixture of discrete distributions with conditional means depending on past observations through periodic time-varying parameters. The mixture feature allows for finitely many component specifications corresponding to the conditional mean and is specified via a finite independent and periodically distributed chain called the switching (or regime) process. Moreover, the lagged values of the conditional mean in each regime are governed by the past values of the regime sequence (cf. [4,9,26,34,40]). Finally, the conditional mean in each regime is a general (linear or nonlinear) function of past observations. The model thus allows for a large spectrum of distributions and various conditional mean shapes.

    The rest of this paper has the following structure. Section 2 defines the model and the tools needed for the subsequent Sections. In Section 3, we propose periodic ergodicity conditions for the mixture periodic INGARCH (henceforth MP-INGARCH). Section 4 extends the results to the case of nonlinear Lipschitz conditional mean forms. Section 5 illustrates the general results on specific mixture periodic distributions such as the periodic Poisson mixture INGARCH, and the periodic negative binomial mixture INGARCH. In addition, a simulation study is carried out to compare the theoretical seasonal expectations generated by the model and their sample counterparts computed using simulated series. Main conclusions are indicated in Section 6.

    In the sequel, it is assumed that all random variables and sequences are defined on a probability space (Ω,F,P) with values in subsets of the integer set N={0,1,...}.

    Let us recall some probability properties for periodic processes such as periodic stationarity and periodic ergodicity. A sequence of random variables {Xt,tZ} is said to be independent and S-periodically distributed (ipdS) if: i) {Xt,tZ} is independent, and ii) Xtd=Xt+kS for all k,tZ, where S1 is a positive integer called the period, and the symbol "d=" stands for equality in distribution. When S=1, an ipd1 sequence is just an independent and identically distributed (iid) process.

    A stochastic process {Yt,tZ} is said to be strictly periodically stationary with period S1 if all processes {YnS+s,nZ} (1sS) are strictly stationary in the ordinary sense. In addition, {Yt,tZ} is called S-periodically stationary in mean if E(Yt) is finite (for all tZ) and is S-periodic over t. Naturally, strict periodic stationarity implies periodic stationarity in the mean whenever the seasonal means of the process are finite. The process {Yt,tZ} is said to be periodically ergodic with period S if all sub-processes {YnS+s,nZ} (s{1,...,S}) are ergodic in the usual sense (cf. [6,17]). The simplest strictly periodically stationary and periodically ergodic process (with period S) is an ipdS sequence.

    Let Fλ be a cumulative distribution function (cdf) with discrete support and mean λ=+0xdFλ(x)>0. Suppose that Fλ satisfies the following "equal stochastic and mean orders" property introduced by [3]

    λλFλ(u)Fλ(u),u(0,1), (2.1)

    where Fλ is the generalized inverse of Fλ. The family of distributions satisfying (2.1) is quite large and contains the one-parameter exponential family such as the Poisson and negative binomial distributions, and also other interesting distributions (cf. [3]). In [11] proposed a large class of periodic INGARCH(p,q) models described as follows. An integer-valued process {Yt,tZ} is said to be a periodic INGARCH(p,q) of orders p and q (p,qN), and positive integer period S1, if its conditional distributions are given by

    Yt|Ft1Ft,λt,tZ, (2.2)

    where Ft1 is the σ-algebra generated by {Yt1,Yt2,...}, the cdf Ft,λt:=Fλt satisfies (2.1) for all tZ, and the conditional mean λt is given by

    λt=ωt+qi=1αtiYti+pj=1βtjλtj,tZ. (2.3)

    The parameters ωt>0, αti0 (i=1,...,q) and βtj0 (j=1,...,p) are periodic in t with period S in the sense ωt=ωt+kS, αti=αt+kS,i and βtj=βt+kS,j for all k,tZ. In a more compact form, Eq (2.3) can be written as follows

    λnS+s=ωs+qi=1αsiYnS+si+pj=1βsjλnS+sj,nZ,1sS.

    In this paper we consider a finite mixture generalization of the periodic INGARCH(p,q) model (2.2)–(2.3). Let a positive integer L, which refers to the number of regimes (or components). Let also {Δt,tZ} be an ipdS sequence of random variables, valued in the finite set {1,...,L} with distribution P(Δt=l)=πt(l), where πt(l)0, Ll=1πt(l)=1, and πt(l) is S-periodic in t in the sense πt+hS(l)=πt(l) for all t, h and l. The values assumed by Δt are called components or regimes and the probability πt(l) is called the mixing proportion along the season or channel t.

    A stochastic process {Yt,tZ} is said to be a mixture periodic INGARCH(p,q) (henceforth MP-INGARCH(p,q)) model if its conditional distribution is a finite mixture of L mixing distributions, that is

    Yt|Ft1πt(1)Ft,1,λ1,t++πt(L)Ft,L,λL,t, (2.4)

    where Fλ=Ft,l,λ satisfies (2.1) and λt:=λΔt,t is given by

    λt=ωt(Δt)+qi=1αti(Δt)Yti+pj=1βtj(Δt)λtj. (2.5)

    For all l, the parameters ωt(l)>0, αti(l)0 and βtj(l)0 are S -periodic over t in the above sense. Equation (2.5) can be written in the periodic form

    λnS+s=ωs(ΔnS+s)+qi=1αsi(ΔnS+s)YnS+si+pj=1βsj(ΔnS+s)λnS+sj,nZ,1sS,

    where, for instance, ωs(l) (1sS, 1lL) denotes the intercept at season s and regime l. Naturally, when L=1, the MP-INGARCH model given by (2.4)–(2.5) reduces to the periodic INGARCH model defined by (2.2)–(2.3) (cf. [11]).

    For identifiability purposes, one can assume that for all 1vS, πv(1)πv(2)πv(L). See also [2,34,40] for time-invariant mixtures corresponding to S=1. Note, however, that the identifiability of the mixture INGARCH model is rather important for asymptotic estimation theory but is not required from a probabilistic point of view.

    It is worth noting that the past recent values of λΔt,t in (2.5) depend on the past values of the regime variable Δt (see also [4,9,24,34]). The likelihood of (2.4)–(2.5) is therefore not simple to obtain as it depends on the whole path information concerning Δt. A more general specification is

    λt=gt,Δt(Yt1,,Ytq,λt1,,λtp), (2.6)

    where gt,l (1lL) is [0,)-valued, and is S -periodic over t.

    Denote by FΔt the sigma-field generated by {Yi,Δi+1,it}. The distribution given by (2.4) can be rewritten in function of Δt as follows

    YtFΔt1Ft,Δt,λt. (2.7)

    Model (2.4)–(2.5) encompasses several important mixtures of distributions such as:

    i) The Poisson mixture

    πt(1)P(λ1,t)++πt(L)P(λL,t). (2.8)

    ii) The quadratic negative binomial mixture (Aknouche and Francq, 2022)

    πt(1)NB(rt(1),rt(1)rt(1)+λ1,t)++πt(L)NB(rt(L),rt(L)rt(L)+λL,t), (2.9)

    where rt(L)>0 (l=1,...,L) is S-periodic over t.

    iii) The linear negative binomial mixture

    πt(1)NB(rt(1)λ1,t,rt(1)rt(1)+1)++πt(L)NB(rt(L)λL,t,rt(L)rt(L)+1). (2.10)

    iv) The mixture of different distributions such as, for example with L=3,

    πt(1)P(λ1t)+πt(2)NB(rt(2),rt(2)rt(2)+λ2,t)+πt(3)NB(rt(3)λ3,t,rt(3)rt(3)+1).

    Conditions under which the MP-INGARCH process defined by (2.4) and (2.5) is strictly periodically stationary and periodically ergodic are now given. Let m=max(p,q) and set

    cti=Ll=1πt(l)(αti(l)+βti(l)),i=1,...,m.

    Consider the m×m companion matrix At given by

    At=(ct1ct2ct,m1ctm100001000010), (3.1)

    Let ρ(B) denote the spectral radius of the matrix B, i.e., the maximum absolute eigenvalues of B.

    Theorem 3.1. Under

    ρ(Ss=1ASs+1)<1, (3.2)

    there exists a strictly periodically stationary, and periodically ergodic process {Yt,tZ} satisfying

    P(YtyFΔt1)=Ft,Δt,λt(y),   yR, (3.3)

    where Ft,Δt,λt fulfills (2.1) for all tZ, and λt satisfies (2.5).

    Conversely, if there is a mean periodically stationary process {Yt,tZ} satisfying (3.3) with E(Yt)< for all tZ, then (3.2) holds.

    Proof. Let us prove the necessary part of the Theorem. If there exists a mean periodically stationary process {Yt,tZ} satisfying (3.2) with E(Yt)=E(λt) for all t, then we have

    E(Yt)=Ll=1πt(l)ωt(l)+mi=1Ll=1πt(l)(αti(l)+βtj(l))E(Yti). (3.4)

    Setting Y_t=(E(Yt),...,E(Ytm+1)) and Bt=(Ll=1πt(l)ωt(l),0,...,0), equality (3.4) can be rewritten in the following matrix equation

    Y_t=AtY_t1+Bt.

    By iterating the latter equation S times while using the S-periodicity in-mean of the process, which implies that E(Y_tS)=E(Y_t), we get the following equation

    Y_t=(AtAtS+1)Y_t+Ct, (3.5)

    where

    Ct=S1j=0j1i=0AtiBtj.

    Using Lemma A.1 in [4] and Corollary 8.1.29 in [31], equality (3.5) entails

    ρ(AtAtS+1)<1,

    which in turns is equivalent to (3.2).

    We now prove the sufficiency part of the theorem. Let {Ut,tZ} be an iid uniformly distributed sequence in [0,1], independent of {Δt,tZ}. For all tZ put

    λ(k)t={ωt(Δt)+qi=1αti(Δt)Y(ki)ti+pj=1βtj(Δt)λ(kj)tj  ifk10                                                             ifk0, (3.6)

    and

    Y(k)t={Ft,Δt,λ(k)t(Ut)  ifk10              ifk0. (3.7)

    If k2, combining (3.6) and (3.7) we have

    λ(k)t=φt,k(Ut1,,Utk+1,Δt,...,Δtk+1),

    where in view of the S-periodicity of the model parameters, the measurable function φt,k:[0,1]k×{1,...,L}k[0,) is S-periodic in t in the sense φt,k=φhS+t,k for all t,hZ. Therefore, the processes {λ(k)t,tZ} and {Y(k)t,tZ} are strictly periodically stationary and periodically ergodic (e.g., [6]) for all k. Now, let F(k)t1 and Ft1 be the σ-algebras generated by {Y(ki)ti,Δti+1,i>0} and {Ui,Δi+1,i<t}, respectively. Since the variable Fλ(U) has the cdf Fλ when U is uniformly distributed in [0,1], it follows that

    E(Y(k)tF(k)t1)=E(Y(k)tFt1)=λ(k)tP(Y(k)tyF(k)t1)=P(Ft,Δt,λ(k)t(Ut)yFt1)=Ft,Δt,λ(k)t(y).

    If we establish the following limiting result

    λt=limkλ(k)t,a.s., (3.8)

    then the existence of a process satisfying (3.3), with Ft1 in place of FΔt1, is proved. Taking the limit as k on the two sides of the equalities (3.6) and (3.7), we obtain a.s.

    Yt=limkY(k)t=Ft,Δt,λt(Ut).

    As λt is FΔt1-measurable, the distribution of Yt given Ft1 is the same as that of Yt given FΔt1. To show that (3.2) implies (3.8), let us first prove that the sequence (λ(k)t)k is increasing, i.e.,

    0λ(k1)tλ(k)t,forallk (3.9)

    and that

    E(Y(k)t)E(Y(k1)t),forallk. (3.10)

    When k0, the inequalities (3.9) and (3.10) are obviously satisfied. For k1, (3.9) is shown by induction. If

    ...λ(k01)tλ(k0)t

    is satisfied then from (2.1) we obtain

    λ(k0+1)t=ωt(Δt)+qi=1αti(Δt)Fλ(k0+1i)ti(Uti)+pj=1βtj(Δt)λ(k0+1j)tjωt(Δt)+qi=1αti(Δt)Fλ(k0i)ti(Uti)+pj=1βtj(Δt)λ(k0j)tj=λ(k0)t.

    Since E(Y(k)t)=E(λ(k)t) exists for any fixed k, we have

    E(Y(k+1)t)=E(λ(k+1)t)E(λ(k)t)=E(Y(k)t),

    which establishes (3.10). Using (3.9) and (3.10), we have for all k1

    E(|λ(k)tλ(k1)t|)mi=1Ll=1πt(l)(αti(l)+βti(l))E(|λ(ki)tiλ(ki1)ti|). (3.11)

    Letting v(k)t=E(|λ(k)tλ(k1)t|) and V(k)t=(v(k)t,...,v(km+1)tm+1), inequality (3.11) can be written in the following matrix form

    V(k)tAtV(k1)t1(AtAtS+1)V(kS)tS=(AtAtS+1)V(kS)t, (3.12)

    where At is given by (3.1), and the equality V(kS)tS=V(kS)t, which follows from the S -periodicity of the coefficients, is used. Under (3.2) and (3.12), we get for all t

    V(k)t0 as k.

    Thus the sequence {λ(k)t}k converges in L1 and a.s. In addition, since

    λt=limkφt,k(Ut1,,Utk+1,Δt,,...Δtk+1):=φt(Ut1,Ut2,;Δt,Δt1,...),

    where φt:[0,1]×{1,...,L}[0,) is measurable and S-periodic in t, the sequence {λt,tZ} is therefore strictly stationary and ergodic in the periodic meaning.

    In view of (3.5) and the S-periodicity of the model parameters, the S unconditional (seasonal) means of the MP-INGARCHS(p,q) model are given, under the periodic ergodicity condition (3.2), by

    {μ_v=(ImSs=1ASs+1)1S1j=0j1i=0AviBvjE(Yv)=(1,0,...,0)1×mμ_v, v{1,...,S}, (3.13)

    where Im stands for the identity matrix of order m=max(p,q).

    This Section extends the periodic ergodicity conditions for the MP-INGARCH(p,q) model when the conditional mean λt has the more general nonlinear form (2.6). Suppose that for all tZ, the S-periodic function gt,l(y1,,yq,λ1,,λp) is Lipschitz, that is for all (yi,yi), i=1,,q and all (λj,λj), j=1,,p,

    |gt,l(y1,,yq,λ1,,λp)gt,l(y1,,yq,λ1,,λp)|qi=1αti(l)|yiyi|+pj=1βtj(l)|λjλj|, (4.1)

    where αti(l)0 and βtj(l)0 are S-periodic over t. See also [34] and [4] for the non-periodic INGARCH setting. Let At be defined as in (3.1). The following result shows that condition (3.3) in which the coefficients of the matrix At are replaced by those of the Lipschitz inequality (4.1), is still sufficient for the existence of a periodically ergodic process verifying (2.7) and (2.6) (or equivalently (2.4) and (2.6))

    Theorem 4.1. There exists a strictly periodically stationary and periodically ergodic process {Yt,tZ} with a conditional distribution given by (2.1) and (2.7), where λt satisfies (2.6) and (4.1), if

    ρ(AtAt1AtS+1)<1. (4.2)

    Proof. Using similar arguments in the proof of Theorem 3.1, let {Ut,tZ} be defined as above. Define for all tZ

    λ(k)t={gt,Δt(Y(k1)t1,,Y(kq)tq,λ(k1)t1,,λ(kp)tp)  ifk10                                                      ifk0 (4.3)

    and let Y(k)t be given as in (3.7). As for the proof of Theorem 3.1, under (4.2) we show the existence of a periodically ergodic process satisfying (2.1), (2.7), (2.6), and (4.1). This amounts to show the a.s. convergence of λ(k)t given by (4.3) to the limit λt, which is given by (2.6). From (2.7) we have

    E(|Y(k)tY(k1)t|)=EE(|Y(k)tY(k1)t||λ(k)t,λ(k1)t)=E(|λ(k)tλ(k1)t|).

    From (4.1) it follows that

    E|λ(k)tλ(k1)t|mi=1Ll=1πt(l)(αti(l)+βti(l))E|λ(ki)tiλ(ki1)ti|k1.

    The latter can be rewritten in the following inequality

    V(k)tAtAtS+1V(kS)t,

    where V(k)t is defined as in (3.12). Under (4.2), V(k)t0 from which the Theorem is established.

    In this subsection, we apply the very general results given by Theorem 3.1 to various special cases.

    Example 5.1. (First-order Poisson MP-INGARCH(1,1) model)

    For model (2.8) and (2.5), taking p=q=1, condition (3.2) becomes

    Ss=1Lk=1πt(k)(αts,1(k)+βts,1(k))<1.

    If in addition L=1, the latter reduces to the periodic ergodicity condition given by [7] for the first-order periodic INGARCH(1,1) model with a Poisson mixture conditional distribution. See also [13,33,35].

    Example 5.2. (First-order Negative binomial MP-INGARCH(1,1) model)

    The condition (3.2) also applies to a vaster class of distributions given by (2.1) such as the negative binomial mixture. We consider here two instances:

    - The quadratic negative binomial mixture ([5]) given by (2.9).

    - The linear negative binomial mixture given by (2.10).

    Example 5.3. (Non-mixed periodic INGARCH(p,q) model)

    When L=1, the matrix At in condition (3.2) reduces to

    At=(αt1+βt1αt2+βt2αt,m1+βt,m1αtm+βtm100001000010),

    which corresponds to the same periodic ergodicity condition given by [11].

    Example 5.4. (The iid mixture INGARCH(p,q) model)

    When S=1 the coefficient cti in the matrix given by (3.1) becomes time-invariant, i.e.,

    ci=Ll=1π(l)(αi(l)+βi(l)),i=1,...,m,

    and therefore condition (3.2) reduces to

    Ll=1π(l)mi=1(αi(l)+βi(l))<1,

    which is the condition (3.6) given by [4].

    Example 5.5. (The standard INGARCH(p,q) model)

    When S=L=1, condition (3.2) reduces to

    ρ(A)<1, (5.1)

    where

    A=(α1+β1α2+β2αm1+βm1αm+βm100001000010),

    and αi (i=1,...,q) and βj (j=1,...,p) are the conditional mean parameters of the time-invariant INGARCH(p,q) model corresponding to S=1. According to Corollary 2.2 in [27], the condition (5.1) is equivalent to

    qi=1αi+pj=1βj<1,

    which is the standard ergodicity condition given by [4].

    Example 5.6. (Mixture periodic INARCH(q) models)

    When p=0 in the general model (2.4)–(2.5), we obtain the particular mixture periodic INARCH model considered by [35]. So our condition (3.2) reduces to their periodic stationarity-in-mean condition. Note, however, that condition (3.2) also ensures the periodic ergodicity of the model, which has not been studied by [35]. Moreover, our condition also applies to larger classes of distributions, to general GARCH lags, and to non-linear conditional mean forms.

    In this Subsection, the veracity of the unconditional mean formula (3.13) (which is based on condition (3.2) of Theorem 3.1) is assessed through a simulation study. Three mixture periodic distributions are considered.

    i) The first one is the two-component Poisson MP-INGARCH4(1,1) model with period S=4,

    Yt|Ft1πt(1)P(λ1,t)+πt(2)P(λ2,t),

    the parameters of which are reported in Table 1.

    Table 1.  Parameters of the Poisson 2-mixture 4-Periodic INGARCH_4(1, 1) model.
    l 1 2 E(Yv)
    v ωv(1) αv(1) βv(1) πv(1) ωv(2) αv(2) βv(2) πv(2)
    1 ω1(1) α1(1) β1(1) π1(1) ω1(2) α1(2) β1(2) π1(2)
    1 0.1 0.5 0.3 1.5 0.2 0.3 0.7 7.200
    2 ω2(1) α2(1) β2(1) π2(1) ω2(2) α2(2) β2(2) π2(2)
    0.5 0.2 0.3 0.7 0.7 0.1 0.5 0.3 4.499
    3 ω3(1) α3(1) β3(1) π3(1) ω3(2) α3(2) β3(2) π3(2)
    3 0.3 0.1 0.4 4 0.4 0.4 0.6 6.497
    4 ω4(1) α4(1) β4(1) π4(1) ω4(2) α4(2) β4(2) π4(2)
    5 0.4 0.2 0.2 6 0.3 0.6 0.8 11.208

     | Show Table
    DownLoad: CSV

    ii) The second one is the two-component (quadratic) negative binomial MP-INGARCH4(1,1) model with period S=4:

    Yt|Ft1πt(1)NB(rt(1),rt(1)rt(1)+λ1,t)+πt(2)NB(rt(2),rt(2)rt(2)+λ2,t),

    where the corresponding parameters are shown in Table 2.

    Table 2.  Parameters of the 2-component negative binomial MP-INGARCH_4(1, 1) model.
    l 1 2 E(Yv)
    v ωv(1) αv(1) βv(1) πv(1) ωv(2) αv(2) βv(2) πv(2)
    rv(1) rv(2)
    1 ω1(1) α1(1) β1(1) π1(1) ω1(2) α1(2) β1(2) π1(2)
    0.50 0.20 0.30 0.40 0.80 0.10 0.50 0.60 8.909
    r1(1) 2.0 r1(2) 1.0
    2 ω2(1) α2(1) β2(1) π2(1) ω2(2) α2(2) β2(2) π2(2)
    7.0 0.10 0.40 0.60 8.0 0.35 0.20 0.40 12.353
    r2(1) 4.0 r2(2) 2.0
    3 ω3(1) α3(1) β3(1) π3(1) ω3(2) α3(2) β3(2) π3(2)
    12.0 0.30 0.20 0.60 10.0 0.45 0.30 0.40 18.512
    r3(1) 0.5 r3(2) 1.0
    4 ω4(1) α4(1) β4(1) π4(1) ω4(2) α4(2) β4(2) π4(2)
    2.0 0.10 0.60 0.80 3.0 0.40 0.20 0.20 14.794
    r4(1) 15.0 r4(2) 10.0

     | Show Table
    DownLoad: CSV

    iii) The third case is the three-component Poisson MP-INARCH4(1) model (p=0) with period S=4 (cf. Table 3), given by

    Yt|Ft1πt(1)P(λ1,t)+πt(2)P(λ2,t)+πt(3)P(λ3,t).
    Table 3.  Parameters of the Poisson 3-mixture 4-Periodic INARCH_4(1, 1) model.
    l 1 2 3 E(Yv)
    v ωv(1) αv(1) πv(1) ωv(2) αv(2) πv(2) ωv(3) αv(3) πv(3)
    1 ω1(1) α1(1) π1(1) ω1(2) α1(2) π1(2) ω1(3) α1(3) π1(3)
    0.5 0.2 0.4 0.8 0.1 0.3 0.7 0.1 0.3 9.113
    2 ω2(1) α2(1) π2(1) ω2(2) α2(2) π2(2) ω2(3) α2(3) π2(3)
    7.0 0.1 0.2 8.0 0.35 0.2 9.0 0.1 0.6 12.439
    3 ω3(1) α3(1) π3(1) ω3(2) α3(2) π3(2) ω3(3) α3(3) π3(3)
    12.0 0.3 0.3 10.0 0.45 0.4 11.0 0.2 0.3 18.775
    4 ω4(1) α4(1) π4(1) ω4(2) α4(2) π4(2) ω4(3) α4(3) π4(3)
    2.0 0.1 0.2 3.0 0.4 0.5 4.0 0.1 0.3 14.848

     | Show Table
    DownLoad: CSV

    For each case, 1000 Monte Carlo replications with sample-size T=800 (hence N=200) are simulated, from which the S seasonal sample means ¯Yv=1NN1n=0YnS+v, for all 1vS, are obtained. Then, these seasonal sample means are compared with their seasonal theoretical counterparts obtained from (3.13) under the periodic stationarity condition (3.2). For the first case, Figure 1 shows the Boxplots of the 4 seasonal sample means and the corresponding theoretical means in solid line. Figures 2 and 3 show the same for cases 2 and 3, respectively.

    Figure 1.  Theoretical and sample seasonal means for the 2-component, 4-periodic Poisson MP-INGARCH4(1,1) model.
    Figure 2.  Theoretical and sample seasonal means for the 2-component, 4-periodic negative binomial MP-INGARCH4(1,1) model.
    Figure 3.  Theoretical and sample seasonal means for the 3-component, 4-periodic Poisson MP-INGARCH4(1,1) model.

    From Figures 13, it can be seen that the theoretical seasonal means are at the center of their corresponding Boxplots, which supports the above theoretical calculations. Moreover, the Boxplots are, in most cases, symmetric and are therefore consistent with the normality assumption, which follows from the central limit theorem for periodically ergodic processes.

    In this paper, we examined some probability properties of a general mixture periodic integer-valued GARCH model, which can be used to model seasonally-varying integer-valued time series data. More precisely, we proposed strict periodic stationarity (and periodic ergodicity) conditions for a wide family of mixture distributions whose stochastic order is the same as the mean order (cf. [4]). The regime sequence driving the mixture feature is assumed to be independent and periodically distributed. Moreover, the lagged conditional means are governed by the lagged values of the mixture variable, which makes the model depends on the past of the regime variable. Therefore, the likelihood of the model is not easy to calculate, but the estimation of the latter can be done using the generalized method of moments ([28]), Bayesian MCMC methods ([12]), or particle filtering ([36]). On the other hand, the model we studied is also general from the form of the conditional means which can be linear or Lipschitz nonlinear.

    Various extensions of the proposed model can be proposed. We mention in particular multivariate mixture periodic models and non-Lipschitz conditional functions such as threshold models.

    I am deeply grateful to the Editor and the referees for their helpful comments and useful suggestions which led to a significant improvement of the earlier version of the paper. The researcher would like to thank the Deanship of Scientific Research, Qassim University, for funding the publication of this project.

    The author declares no conflicts of interest regarding this article.



    [1] A. Acker, Interior free boundary problems for the Laplace equation, Arch. Rational Mech. Anal., 75 (1981), 157–168. https://doi.org/10.1007/BF00250477 doi: 10.1007/BF00250477
    [2] A. Acker, Uniqueness and monotonicity of solutions for the interior Bernoulli free boundary problem in the convex, n-dimensional case, Nonlinear Anal. Theor., 13 (1989), 1409–1425. https://doi.org/10.1016/0362-546X(89)90102-8 doi: 10.1016/0362-546X(89)90102-8
    [3] A. Aftalion, J. Busca, W. Reichel, Approximate radial symmetry for overdetermined boundary value problems, Adv. Differential Equations, 4 (1999), 907–932.
    [4] V. Agostiniani, R. Magnanini, Symmetries in an overdetermined problem for the Green's function, Discrete Cont. Dyn. Syst. S, 4 (2011), 791–800. https://doi.org/10.3934/dcdss.2011.4.791 doi: 10.3934/dcdss.2011.4.791
    [5] C. Bianchini, A Bernoulli problem with non-constant gradient boundary constraint, Appl. Anal., 91 (2012), 517–527. https://doi.org/10.1080/00036811.2010.549479 doi: 10.1080/00036811.2010.549479
    [6] C. Bianchini, A. Henrot, P. Salani, An overdetermined problem with non-constant boundary condition, Interfaces Free Bound., 16 (2014), 215–241. https://doi.org/10.4171/IFB/318 doi: 10.4171/IFB/318
    [7] B. Brandolini, C. Nitsch, P. Salani, C. Trombetti, Serrin-type overdetermined problems: allternative proof, Arch. Rational Mech. Anal., 190 (2008), 267–280. https://doi.org/10.1007/s00205-008-0119-3 doi: 10.1007/s00205-008-0119-3
    [8] B. Brandolini, C. Nitsch, P. Salani, C. Trombetti, On the stability of the Serrin problem, J. Differ. Equations, 245 (2008), 1566–1583. https://doi.org/10.1016/j.jde.2008.06.010 doi: 10.1016/j.jde.2008.06.010
    [9] L. A. Caffarelli, H. W. Alt, Existence and regularity for a minimum problem with free boundary, J. Reine Angew. Math., 325 (1981), 105–144. https://doi.org/10.1515/crll.1981.325.105 doi: 10.1515/crll.1981.325.105
    [10] L. Cavallina, T. Yachimura, On a two-phase Serrin-type problem and its numerical computation, ESAIM: COCV, 26 (2020), 65. https://doi.org/10.1051/cocv/2019048 doi: 10.1051/cocv/2019048
    [11] L. Cavallina, Local analysis of a two phase free boundary problem concerning mean curvature, Indiana Univ. Math. J., in press.
    [12] L. Cavallina, T. Yachimura, Symmetry breaking solutions for a two-phase overdetermined problem of Serrin-type, In: Current trends in analysis, its applications and computation (Proceedings of the 12th ISAAC Congress, Aveiro, Portugal, 2019), Birkhäuser, 2022, in press.
    [13] G. Ciraolo, R. Magnanini, V. Vespri, Hölder stability for Serrin's overdetermined problem, Annali di Matematica, 195 (2016), 1333–1345. https://doi.org/10.1007/s10231-015-0518-7 doi: 10.1007/s10231-015-0518-7
    [14] M. Del Pino, F. Pacard, J. Wei, Serrin's overdetermined problem and constant mean curvature surfaces, Duke Math. J., 164 (2015), 2643–2722. https://doi.org/10.1215/00127094-3146710 doi: 10.1215/00127094-3146710
    [15] M. M. Fall, I. A. Minlend, T. Weth, Unbounded periodic solutions to Serrin's overdetermined boundary value problem, Arch. Rational Mech. Anal., 223 (2017), 737–759. https://doi.org/10.1007/s00205-016-1044-5 doi: 10.1007/s00205-016-1044-5
    [16] W. M. Feldman, Stability of Serrin's problem and dynamic stability of a model for contact angle motion, SIAM J. Math. Anal., 50 (2018), 3303–3326. https://doi.org/10.1137/17M1143009 doi: 10.1137/17M1143009
    [17] D. Gilbarg, N. S. Trudinger, Elliptic partial differential equations of second order, Berlin, Heidelberg: Springer, 2001. https://doi.org/10.1007/978-3-642-61798-0
    [18] A. Gilsbach, M. Onodera, Linear stability estimates for Serrin's problem via a modified implicit function theorem, Calc. Var., 60 (2021), 241. https://doi.org/10.1007/s00526-021-02107-1 doi: 10.1007/s00526-021-02107-1
    [19] A. Henrot, M. Onodera, Hyperbolic solutions to Bernoulli's free boundary problem, Arch. Rational Mech. Anal., 240 (2021), 761–784. https://doi.org/10.1007/s00205-021-01620-z doi: 10.1007/s00205-021-01620-z
    [20] A. Henrot, H. Shahgholian, The one phase free boundary problem for the p-Laplacian with non-constant Bernoulli boundary condition, Trans. Amer. Math. Soc., 354 (2002), 2399–2416. https://doi.org/10.1090/S0002-9947-02-02892-1 doi: 10.1090/S0002-9947-02-02892-1
    [21] N. Kamburov, L. Sciaraffia, Nontrivial solutions to Serrin's problem in annular domains, Ann. Inst. H. Poincaré C Anal. Non Linéaire, 38 (2021), 1–22. https://doi.org/10.1016/j.anihpc.2020.05.001 doi: 10.1016/j.anihpc.2020.05.001
    [22] R. Magnanini, G. Poggesi, On the stability for Alexandrov's soap bubble theorem, J. Anal. Math., 139 (2019), 179–205. https://doi.org/10.1007/s11854-019-0058-y doi: 10.1007/s11854-019-0058-y
    [23] R. Magnanini, G. Poggesi, Serrin's problem and Alexandrov's soap bubble theorem: enhanced stability via integral identities, Indiana Univ. Math. J., 69 (2020), 1181–1205. https://doi.org/10.1512/iumj.2020.69.7925 doi: 10.1512/iumj.2020.69.7925
    [24] R. Magnanini, G. Poggesi, Nearly optimal stability for Serrin's problem and the Soap Bubble theorem, Calc. Var., 59 (2020), 35. https://doi.org/10.1007/s00526-019-1689-7 doi: 10.1007/s00526-019-1689-7
    [25] R. Magnanini, G. Poggesi, Interpolating estimates with applications to some quantitative symmetry results, Mathematics in Engineering, 5 (2023), 1–21. https://doi.org/10.3934/mine.2023002 doi: 10.3934/mine.2023002
    [26] F. Morabito, P. Sicbaldi, Delaunay type domains for an overdetermined elliptic problem in Sn×R and Hn×R, ESAIM: COCV, 22 (2016), 1–28. https://doi.org/10.1051/cocv/2014064 doi: 10.1051/cocv/2014064
    [27] F. Morabito, Symmetry breaking bifurcations for two overdetermined boundary value problems with non-constant Neumann condition on exterior domains in R3, Commun. Part. Diff. Eq., 46 (2021), 1137–1161. https://doi.org/10.1080/03605302.2020.1871363 doi: 10.1080/03605302.2020.1871363
    [28] M. Onodera, Geometric flows for quadrature identities, Math. Ann., 361 (2015), 77–106. https://doi.org/10.1007/s00208-014-1062-2 doi: 10.1007/s00208-014-1062-2
    [29] M. Onodera, On the symmetry in a heterogeneous overdetermined problem, Bull. Lond. Math. Soc., 47 (2015), 95–100. https://doi.org/10.1112/blms/bdu098 doi: 10.1112/blms/bdu098
    [30] M. Onodera, Dynamical approach to an overdetermined problem in potential theory, J. Math. Pure. Appl., 106 (2016), 768–796. https://doi.org/10.1016/j.matpur.2016.03.011 doi: 10.1016/j.matpur.2016.03.011
    [31] L. E. Payne, P. W. Schaefer, Duality theorems in some overdetermined boundary value problems, Math. Method. Appl. Sci., 11 (1989), 805–819. https://doi.org/10.1002/mma.1670110606 doi: 10.1002/mma.1670110606
    [32] G. Pólya, Torsional rigidity, principal frequency, electrostatic capacity and symmetrization, Quart. Appl. Math., 6 (1948), 267–277. https://doi.org/10.1090/QAM/26817 doi: 10.1090/QAM/26817
    [33] F. Schlenk, P. Sicbaldi, Bifurcating extremal domains for the first eigenvalue of the Laplacian, Adv. Math., 229 (2012), 602–632. https://doi.org/10.1016/j.aim.2011.10.001 doi: 10.1016/j.aim.2011.10.001
    [34] J. Serrin, A symmetry problem in potential theory, Arch. Rational Mech. Anal., 43 (1971), 304–318. https://doi.org/10.1007/BF00250468 doi: 10.1007/BF00250468
    [35] H. Shahgholian, Diversifications of Serrin's and related symmetry problems, Complex Var. Elliptic Equ., 57 (2012), 653–665. https://doi.org/10.1080/17476933.2010.504848 doi: 10.1080/17476933.2010.504848
    [36] P. Sicbaldi, New extremal domains for the first eigenvalue of the Laplacian in flat tori, Calc. Var., 37 (2010), 329–344. https://doi.org/10.1007/s00526-009-0264-z doi: 10.1007/s00526-009-0264-z
    [37] H. F. Weinberger, Remark on the preceding paper of Serrin, Arch. Rational Mech. Anal., 43 (1971), 319–320. https://doi.org/10.1007/BF00250469 doi: 10.1007/BF00250469
  • This article has been cited by:

    1. BADER S. ALMOHAIMEED, A NEGATIVE BINOMIAL AUTOREGRESSION WITH A LINEAR CONDITIONAL VARIANCE-TO-MEAN FUNCTION, 2022, 30, 0218-348X, 10.1142/S0218348X22402393
    2. Bader S. Almohaimeed, Correction: Periodic stationarity conditions for mixture periodic INGARCH models, 2022, 7, 2473-6988, 18280, 10.3934/math.20221005
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

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

Metrics

Article views(2407) PDF downloads(250) Cited by(3)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog