Processing math: 71%
Research article Special Issues

Multifunctional farm advisory services in promoting change in agricultural systems: The case of Campania region of Italy

  • Entrepreneurial contexts may be marked by the presence of a 'cultural environment' that stimulates knowledge and innovation adoption, while other contexts may act as barriers toward change and innovation. Moreover, multiple paths of multifunctional agriculture bring about a call for "multifunctional farm advisory services" (MFAS), which consider both private and public goods provided by the farming sector. Set against the background of multiple roles of agriculture, how to identify sound and pertinent knowledge becomes of paramount, to specify the roles of agricultural extensionists and the mechanisms of governance of MFAS within the setting up of the Agricultural Knowledge and Innovation System (AKIS). Our aim of the study is either to analyze attitudes toward the privatization of extension services within a predominantly public system of regional governance and to identify advisors' profile and their suitability with the modern vision of multifunctional agriculture through the emergence of MFAS. Empirical analysis evidences the presence of a diversified set of advisory services with different degrees of coherence with the multifunctional agricultural model. Also, the more advisory services are oriented towards empowering multifunctional agriculture the less the propensity towards their privatization. The cluster analysis has demonstrated a relatively good advisor's capability to deal with the new demands of multifunctional agriculture. The idea of MFAS has important theoretical implications that the paper tries to excavate through the analysis of the mechanisms of governance (public/private) and the identification of the advisors' profile facing the growing complexity of the farming sector, grounded on multifunctional agriculture. The study tries to fill a gap in the literature, by providing an original contribution to modeling the profile of advisors in charge of supporting the transition towards multifunctionality.

    Citation: Marcello De Rosa, Giuseppina Olivieri, Concetta Menna, Ferdinando Gandolfi, Teresa Del Giudice. Multifunctional farm advisory services in promoting change in agricultural systems: The case of Campania region of Italy[J]. AIMS Agriculture and Food, 2023, 8(4): 962-977. doi: 10.3934/agrfood.2023051

    Related Papers:

    [1] Yan Ling Fu, Wei Zhang . Some results on frames by pre-frame operators in Q-Hilbert spaces. AIMS Mathematics, 2023, 8(12): 28878-28896. doi: 10.3934/math.20231480
    [2] Gang Wang . Some properties of weaving K-frames in n-Hilbert space. AIMS Mathematics, 2024, 9(9): 25438-25456. doi: 10.3934/math.20241242
    [3] Sergio Verdú . Relative information spectra with applications to statistical inference. AIMS Mathematics, 2024, 9(12): 35038-35090. doi: 10.3934/math.20241668
    [4] Ligong Wang . Output statistics, equivocation, and state masking. AIMS Mathematics, 2025, 10(6): 13151-13165. doi: 10.3934/math.2025590
    [5] Cure Arenas Jaffeth, Ferrer Sotelo Kandy, Ferrer Villar Osmin . Functions of bounded (2,k)-variation in 2-normed spaces. AIMS Mathematics, 2024, 9(9): 24166-24183. doi: 10.3934/math.20241175
    [6] Chibueze C. Okeke, Abubakar Adamu, Ratthaprom Promkam, Pongsakorn Sunthrayuth . Two-step inertial method for solving split common null point problem with multiple output sets in Hilbert spaces. AIMS Mathematics, 2023, 8(9): 20201-20222. doi: 10.3934/math.20231030
    [7] Osmin Ferrer Villar, Jesús Domínguez Acosta, Edilberto Arroyo Ortiz . Frames associated with an operator in spaces with an indefinite metric. AIMS Mathematics, 2023, 8(7): 15712-15722. doi: 10.3934/math.2023802
    [8] Abdullah Ali H. Ahmadini, Amal S. Hassan, Ahmed N. Zaky, Shokrya S. Alshqaq . Bayesian inference of dynamic cumulative residual entropy from Pareto Ⅱ distribution with application to COVID-19. AIMS Mathematics, 2021, 6(3): 2196-2216. doi: 10.3934/math.2021133
    [9] Messaoud Bounkhel . V-Moreau envelope of nonconvex functions on smooth Banach spaces. AIMS Mathematics, 2024, 9(10): 28589-28610. doi: 10.3934/math.20241387
    [10] Jamilu Adamu, Kanikar Muangchoo, Abbas Ja'afaru Badakaya, Jewaidu Rilwan . On pursuit-evasion differential game problem in a Hilbert space. AIMS Mathematics, 2020, 5(6): 7467-7479. doi: 10.3934/math.2020478
  • Entrepreneurial contexts may be marked by the presence of a 'cultural environment' that stimulates knowledge and innovation adoption, while other contexts may act as barriers toward change and innovation. Moreover, multiple paths of multifunctional agriculture bring about a call for "multifunctional farm advisory services" (MFAS), which consider both private and public goods provided by the farming sector. Set against the background of multiple roles of agriculture, how to identify sound and pertinent knowledge becomes of paramount, to specify the roles of agricultural extensionists and the mechanisms of governance of MFAS within the setting up of the Agricultural Knowledge and Innovation System (AKIS). Our aim of the study is either to analyze attitudes toward the privatization of extension services within a predominantly public system of regional governance and to identify advisors' profile and their suitability with the modern vision of multifunctional agriculture through the emergence of MFAS. Empirical analysis evidences the presence of a diversified set of advisory services with different degrees of coherence with the multifunctional agricultural model. Also, the more advisory services are oriented towards empowering multifunctional agriculture the less the propensity towards their privatization. The cluster analysis has demonstrated a relatively good advisor's capability to deal with the new demands of multifunctional agriculture. The idea of MFAS has important theoretical implications that the paper tries to excavate through the analysis of the mechanisms of governance (public/private) and the identification of the advisors' profile facing the growing complexity of the farming sector, grounded on multifunctional agriculture. The study tries to fill a gap in the literature, by providing an original contribution to modeling the profile of advisors in charge of supporting the transition towards multifunctionality.



    In [13,14,15,16,19], it was proposed that insight into a probability distribution, μ, posed on a Hilbert space, H, could be obtained by finding a best fit Gaussian approximation, ν. This notion of best, or optimal, was with respect to the relative entropy, or Kullback-Leibler divergence:

    R(ν||μ)={Eν[logdνdμ],νμ,+,otherwise. (1.1)

    Having a Gaussian approximation provides qualitative insight into μ, as it provides a concrete notion of the mean and variance of the distribution. Additionally, this optimized distribution can be used in algorithms, such as random walk Metropolis, as a preconditioned proposal distribution to improve performance. Such a strategy can benefit a number of applications, including path space sampling for molecular dynamics and parameter estimation in statistical inverse problems.

    Observe that in the definition of R, (1.1), there is an asymmetry in the arguments. Were we to work with R(μ||ν), our optimal Gaussian would capture the first and second moments of μ, and in some applications this is desirable. However, for a multimodal problem (consider a distribution with two well separated modes), this would be inadequate; our form attempts to match individual modes of the distribution by a Gaussian. For a recent review of the R(ν||μ) problem, see [4], where it is remarked that this choice of arguments is likely to underestimate the dispersion of the distribution of interest, μ. The other ordering of arguments has been explored, in the finite dimensional case, in [2,3,10,18].

    To be of computational use, it is necessary to have an algorithm that will converge to this optimal distribution. In [15], this was accomplished by first expressing ν=N(m,C(p)), where m is the mean and p is a parameter inducing a well defined covariance operator, and then solving the problem,

    (m,p)argminR(N(m,C(p))||μ), (1.2)

    over an admissible set. The optimization step itself was done using the Robbins-Monro algorithm (RM), [17], by seeking a root of the first variation of the relative entropy. While the numerical results of [15] were satisfactory, being consistent with theoretical expectations, no rigorous justification for the application of RM to the examples was given.

    In this work, we emphasize the study and application of RM to potentially infinite dimensional problems. Indeed, following the framework of [15,16], we assume that μ is posed on the Borel σ-algebra of a separable Hilbert space (H,,,). For simplicity, we will leave the covariance operator C fixed, and only optimize over the mean, m. Even in this case, we are seeking mH, a potentially infinite-dimensional space.

    Given the objective function f:HH, assume that it has a root, x. In our application to relative entropy, f will be its first variation. Further, we assume that we can only observe a noisy version of f, F:H×χH, such that for all xH,

    f(x)=E[F(x,Z)]=χF(x,z)μZ(dz), (1.3)

    where μZ is the distribution associated with the random variable (r.v.) Z, taking values in the auxiliary space χ. The naive Robbins-Monro algorithm is given by

    Xn+1=Xnan+1F(Xn,Zn+1), (1.4)

    where ZnμZ, are independent and identically distributed (i.i.d.), and an>0 is a carefully chosen sequence. Subject to assumptions on f, F, and the distribution μZ, it is known that Xn will converge to x almost surely (a.s.), in finite dimensions, [5,6,17]. Often, one needs to assume that f grows at most linearly,

    f(x)c0+c1x, (1.5)

    in order to apply the results in the aforementioned papers. The analysis in the finite dimensional case has been refined tremendously over the years, including an analysis based on continuous dynamical systems. We refer the reader to the books [1,8,11] and references therein.

    As noted, much of the analysis requires the regression function f to have, at most, linear growth. Alternatively, an a priori assumption is sometimes made that the entire sequence generated by (1.4) stays in a bounded set. Both assumptions are limiting, though, in practice, one may find that the algorithms converge.

    One way of overcoming these assumptions, while still ensuring convergence, is to introduce trust regions that the sequence {Xn} is permitted to explore, along with a "truncation" which enforces the constraint. Such truncations distort (1.4) into

    Xn+1=Xnan+1F(Xn,Zn+1)+an+1Pn+1, (1.6)

    where Pn+1 is the projection keeping the sequence {Xn} within the trust region. Projection algorithms are also discussed in [1,8,11].

    We consider RM on a possibly infinite dimensional separable Hilbert space. This is of particular interest as, in the context of relative entropy optimization, we may be seeking a distribution in a Sobolev space associated with a PDE model. A general analysis of RM with truncations in Hilbert spaces can be found in [20]. The main purpose of this work is to adapt the analysis of [12] to the Hilbert space setting for two versions of the truncated problem. The motivation for this is that the analysis of [12] is quite straightforward, and it is instructive to see how it can be easily adapted to the infinite dimensional setting. The key modification in the proof is that results for Banach space valued martingales must be invoked. We also adapt the results to a version of the algorithm where there is prior knowledge on the location of the root. With these results in hand, we can then verify that the relative entropy minimization problem can be solved using RM.

    In some problems, one may have a priori information on the root. For instance, we may know that xU1, some open bounded set. In this version of the truncated algorithm, we have two open bounded sets, U0U1, and xU1. Let σ0=0 and X0U0 be given, then (1.6) can be formulated as

    ˜Xn+1=Xnan+1F(Xn,Zn+1) (1.7a)
    Xn+1={˜Xn+1˜Xn+1U1X(σn)0˜Xn+1U1 (1.7b)
    σn+1={σn˜Xn+1U1σn+1˜Xn+1U1 (1.7c)

    We interpret ˜Xn+1 as the proposed move, which is either accepted or rejected depending on whether or not it will remain in the trust region. If it is rejected, the algorithm restarts at X(σn)0U0. The restart points, {X(σn)0}, may be random, or it may be that X(σn)0=X0 is fixed. The essential property is that the algorithm will restart in the interior of the trust region, away from its boundary. The r.v. σn counts the number of times a truncation has occurred. Algorithm (1.7) can now be expressed as

    Xn+1=Xnan+1F(Xn,Zn+1)+Pn+1Pn+1={X(σn)0˜Xn+1}1˜Xn+1U1. (1.8)

    In the second version of truncated Robbins-Monro, define the sequence of open bounded sets, Un such that:

    U0U1U2,n=0Un=H. (1.9)

    Again, letting X0U0, σ0=0, the algorithm is

    ˜Xn+1=Xnan+1F(Xn,Zn+1) (1.10a)
    Xn+1={˜Xn+1˜Xn+1UσnX(σn)0˜Xn+1Uσn (1.10b)
    σn+1={σn˜Xn+1Uσnσn+1˜Xn+1Uσn (1.10c)

    A consequence of this formulation is that XnUσn for all n. As before, the restart points may be random or fixed, and they are in U0. This would appear superior to the fixed trust region algorithm, as it does not require knowledge of the sets. However, to guarantee convergence, global (in H) assumptions on the regression function are required; see Assumption 2 below. (1.10) can written with Pn+1 as

    Xn+1=Xnan+1F(Xn,Zn+1)+Pn+1Pn+1={X(σn)0˜Xn+1}1˜Xn+1Uσn (1.11)

    In Section 2, we state sufficient assumptions for which we are able to prove convergence in both the fixed and expanding trust region problems, and we also establish some preliminary results. In Section 3, we focus on the relative entropy minimization problem, and identify what assumptions must hold for convergence to be guaranteed. Examples are then presented in Section 4, and we conclude with remarks in Section 5.

    We first reformulate (1.8) and (1.15) in the more general form

    Xn+1=Xnan+1f(Xn)an+1δMn+1=˜Xn+1+an+1Pn+1, (2.1)

    where δMn+1, the noise term, is

    δMn+1=F(Xn,Zn+1)f(Xn)=F(Xn,Zn+1)E[F(Xn,Zn+1)Xn]. (2.2)

    A natural filtration for this problem is Fn=σ(X0,Z1,,Zn). Xn is Fn measurable and the noise term can be expressed in terms of the filtration as δMn+1=F(Xn,Zn+1)E[F(Xn,Zn+1)Fn].

    We now state our main assumptions:

    Assumption 1. f has a zero, x. In the case of the fixed trust region problem, there exist R0<R1 such that

    U0BR0(x)BR1(x)U1.

    In the case of the expanding trust region problem, the open sets are defined as Un=Brn(0) with

    0<r0<r1<r2<<rn. (2.3)

    These sets clearly satisfy (1.9).

    Assumption 2. For any 0<a<A, there exists δ>0:

    infaxxAxx,f(x)δ.

    In the case of the fixed truncation, this inequality is restricted to xU1. This is akin to a convexity condition on a functional F with f=DF.

    Assumption 3. xE[F(x,Z)2] is bounded on bounded sets, with the restriction to U1 in the case of fixed trust regions.

    Assumption 4. an>0, an=, and a2n<

    Theorem 2.1. Under the above assumptions, for the fixed trust region problem, Xnx a.s. and σn is a.s. finite.

    Theorem 2.2. Under the above assumptions, for the expanding trust region problem, Xnx a.s. and σn is a.s. finite.

    Note the distinction between the assumptions in the two algorithms. In the fixed truncation algorithm, Assumptions 2 and 3 need only hold in the set U1, while in the expanding truncation algorithm, they must hold in all of H. While this would seem to be a weaker condition, it requires identification of the sets U0 and U1 for which the assumptions hold. Such sets may not be readily identifiable, as we will see in our examples.

    We first need some additional information about f and the noise sequence δMn.

    Lemma 2.1. Under Assumption 3, f is bounded on U1, for the fixed trust region problem, and on arbitrary bounded sets, for the expanding trust region problem.

    Proof. Trivially,

    f(x)=E[F(x,Z)]E[F(x,Z)]E[F(x,Z)2],

    and the results follows from the assumption.

    Proposition 2.1. For the fixed trust region problem, let

    Mn=ni=1aiδMi.

    Alternatively, in the expanding trust region problem, for r>0, let

    Mn=ni=1aiδMi1Xi1xr.

    Under Assumptions 3 and 4, Mn is a martingale, converging in H, a.s.

    Proof. The following argument holds in both the fixed and expanding trust region problems, with appropriate modifications. We present the expanding trust region case. The proof is broken up into 3 steps:

    1. Relying on Theorem 6 of [7] for Banach space valued martingales, it will be sufficient to show that Mn is a martingale, uniformly bounded in L1(P).

    2. In the case of the expanding truncations,

    E[δMi1Xi1xr2]2E[F(Xi1,Zi)1Xi1xr2]+2E[f(Xi1)1Xi1xr2]2supxxrE[F(x,Z)2]+2supxxrf(x)2

    Since both of these terms are bounded, independently of i, by Assumption 3 and Lemma 1, this is finite.

    3. Next, since {δMi1Xi1xr} is a martingale difference sequence, we can use the above estimate to obtain the uniform L2(P) bound,

    E[Mn2]=ni=1a2iE[δMi1Xi1xr2]supiE[δMi1Xi1xr2]i=1a2i<

    Uniform boundedness in L2, gives boundedness in L1, and this implies a.s. convergence in H.

    In this section we prove results showing that only finitely many truncations will occur, in either the fixed or expanding trust region case. Recall that when a truncation occurs, the equivalent conditions hold: Pn+10; σn+1=σn+1; and ˜Xn+1U1 in the fixed trust region algorithm, while ˜Xn+1Uσn in the expanding trust region case.

    Lemma 2.2. In the fixed trust region algorithm, if Assumptions 1, 2, 3, and 4 hold, then the number of truncations is a.s. finite; a.s., there exists N, such that for all nN, σn=σN.

    Proof. We break the proof up into 7 steps:

    1. Pick ρ and ρ such that

    R0<ρ<ρ<R1 (2.4)

    Let ˉf=supf(x), with the supremum over U1; this bound exists by Lemma 1. Under Assumption 2, there exists δ>0 such that

    infR0/2xxR1xx,f(x)=δ. (2.5)

    Having fixed ρ, ρ, ˉf, and δ, take ϵ>0 such that:

    ϵ<min{ρR0,R1ρ2+ˉf,ρR0ˉf,R02,δ2ˉf,δˉf2,ρρ}. (2.6)

    Having fixed such an ϵ, by the assumptions of this lemma and Proposition 1, a.s., there exists nϵ such that for any n,mnϵ, both

    mk=nakδMkϵ,anϵ. (2.7)

    2. Define the auxiliary sequence

    Xn=Xnk=n+1akδMk. (2.8)

    Using (2.1), we can then write

    Xn+1=Xnan+1f(Xn)+an+1Pn+1. (2.9)

    By (2.7), for any nnϵ,

    XnXnϵ (2.10)

    3. We will show XnBρ(x) for all n large enough. The significance of this is that if nnϵ, and XnBρ(x), then no truncation occurs. Indeed, using (2.6)

    ˜Xn+1xXnx+XnXn+an+1ˉf+an+1δMn+1<ρ+ϵ+ϵˉf+ϵ<R1,˜Xn+1U1. (2.11)

    Consequently, Pn+1=0, Xn+1=˜Xn+1, and σn+1=σn. Thus, establishing XnBρ(x) will yield the result.

    4. Let

    N=inf{nnϵ˜Xn+1U1}+1 (2.12)

    This corresponds to the the first truncation after nϵ. If the above set is empty, for that realization, no truncations occur after nϵ, and we are done. In such a case, we may take N=nϵ in the statement of the lemma.

    5. We now prove by induction that in the case that (2.12) is finite, XnBρ(x) for all nN. First, note that XNBR0(x)Bρ(x). By (2.6) and (2.10),

    XNxXNx+XNXN<R0+ϵ<ρ,XNBρ(x).

    Next, assume XN,XN+1,,Xn are all in Bρ(x). Using (2.11), we have that PN+1==Pn+1=0 and σN==σn=σn+1. Therefore,

    Xn+1x2=Xnx22an+1Xnx,f(Xn)+a2n+1f(Xn)2Xnx22an+1Xnx,f(Xn)+an+1ϵˉf2 (2.13)

    We now consider two cases of (2.13) to conclude Xn+1x<ρ.

    6. In the first case, XnxR0. By Cauchy-Schwarz and (2.6)

    Xn+1x2<R20+2ϵR0ˉf+ϵ2ˉf2=(R0+ϵˉf)2<(ρ)2.

    In the second case, R0<Xnx<ρ. Dissecting the inner product term in (2.13) and using Assumption 2 and (2.10),

    Xnx,f(Xn)=Xnx,f(Xn)+XnXn,f(Xn)Xnx,f(Xn)ˉfϵ (2.14)

    Conditions (2.6) and (2.10) yield the following upper and lower bounds:

    XnxXnxXnXnR0ϵ>12R0,XnxXnx+XnXnρ+ϵ<ρ<R1.

    Therefore, (2.5) applies and Xnx,f(Xn)δ. Using this in (2.14), and condition (2.6),

    Xnx,f(Xn)δˉfϵ>12δ.

    Substituting this last estimate back into (2.13), and using (2.6),

    Xn+1x2<(ρ)2an+1(δϵˉf2)<(ρ)2.

    This completes the inductive step.

    7. Since the auxiliary sequence remains in Bρ(x) for all nN>nϵ, (2.11) ensures ˜Xn+1BR1(x), Pn+1=0, and σn+1=σN, a.s.

    To obtain a similar result for the expanding trust region problem, we first relate the finiteness of the number of truncations with the sequence persisting in a bounded set.

    Lemma 2.3. In the expanding trust region algorithm, if Assumptions 1, 3, and 4 hold, then the sequence remains in a set of the form BR(0) for some R>0 if and only if the number of truncations is finite, a.s.

    Proof. We break this proof into 4 steps:

    1. If the number of truncations is finite, then there exists N such that for all nN, σn=σN. Consequently, the proposed moves are always accepted, and XnUσn=UσN for all nN. Since XnUσnUσN for n<N, XnUσN for all n. By Assumption 3, BR(0)=BrσN(0)=UσN is the desired set.

    2. For the other direction, assume that there exists R>0 such that XnBR(0) for all n. Since the rn in (2.3) tend to infinity, there exists N1, such that R<R+1<rN1. Hence, for all nN1,

    BR(0)BR+1(0)Un (2.15)

    Let ˉf=supf(x), with the supremum over BR(0). Let ˜R be sufficiently large such that BR+1(0)B˜R(x). Lastly, using Proposition 1 and Assumption 4, a.s., there exists N2, such that for all nN2

    anδMn1Xnx˜R<12,an<12(1+ˉf) (2.16)

    Since XnBR(0)B˜R(x), the indicator function in (2.16) is always one, and anδMn<1/2.

    3. Next, let

    N=inf{n0σnmax{N1,N2}} (2.17)

    If the above set is empty, then σn<max{N1,N2} for all n, and the number of truncations is a.s. finite. In this case, the proof is complete.

    4. If the set in (2.17) is not empy, then N<. Take nN. As XnBR(0), and since nσnmax{N1,N2}, (2.16) applies. Therefore,

    ˜Xn+1Xn+˜Xn+1XnXn+an+1f(Xn)+an+1δMn+1<R+12+12<R+1. (2.18)

    Thus, ˜Xn+1BR+1(0)UN1, σnN1, and UN1Uσn. Therefore, ˜Xn+1Uσn. No truncation occurs, and σn=σn+1. Since this holds for all nN, σn=σN, and the number of truncations is a.s. finite.

    Next, we establish that, subject to an additional assumption, the sequence remains in a bounded set; the finiteness of the truncations is then a corollary.

    Lemma 2.4. In the expanding trust region algorithm, if Assumptions 1, 2, 3, and 4 hold, and for any r>0, there a.s. exists N<, such that for all nN,

    Pn+11Xnxr=0,

    then {Xn} remains in a bounded open set, a.s.

    Proof. We break this proof into 7 steps:

    1. We begin by setting some constants for the rest of the proof. Fix R>0 sufficiently large such that BR(x)U0. Next, let ˉf=supf(x) with the supremum taken over BR+2(x). Assumption 2 ensures there exists δ>0 such that

    infR/2xxR+2xx,f(x)=δ. (2.19)

    Having fixed R, ˉf, and δ, take ϵ>0 such that:

    ϵ<min{1,1ˉf,δ2ˉf,δˉf2,R2}. (2.20)

    By the assumptions of this lemma and Proposition 1 there exists, a.s., nϵN such that for all nnϵ,

    i=n+1aiδMi1Xi1xR+2ϵ, (2.21a)
    Pn+11XnxR+2=0, (2.21b)
    an+1ϵ (2.21c)

    2. Define the modified sequence for nnϵ as

    Xn=Xnk=n+1akδMk1Xk1xR+2,XnXnϵ. (2.22)

    Using (2.1), we have the iteration

    Xn+1=Xnan+1δMn+11Xnx>R+2an+1f(Xn)+an+1Pn+1. (2.23)

    3. Let

    N=inf{nnϵσn+1σn}+1, (2.24)

    the first time after nϵ that a truncation occurs.

    If the above set is empty, no truncations occur after nϵ. In this case, σn=σnϵnϵ< for all nnϵ. Therefore, for all nnϵ, XnUσnUσnϵ. Since UσnUσnϵ for all n<nϵ too, the proof is complete in this case.

    4. Now assume that N<. We will show that {Xn} remains in BR+1(x) for all nN. Were this to hold, then for nN,

    XnxXnx+i=n+1aiδMi1Xi1xR+2<R+1+ϵ<R+2, (2.25)

    having used (2.21) and (2.22). For n<N, XnUσnUσN=BrN(0). Therefore, for all n, XnB˜R(0) where ˜R=max{rN,x+R+2}.

    5. We prove XnBR+1(x) by induction. First, since ϵ<1 and XNU0BR(x),

    XNxXNXN+XNx<ϵ+R<R+1.

    Next, assume that XN,XN+1,,Xn are all in BR+1(x). By (2.25), XnBR+2(x). Since Pn+11XnxR+2=0, we conclude Pn+1=0. The modified iteration (2.23) simplifies to have

    Xn+1=Xnan+1f(Xn),

    and

    Xn+1x2=Xnx22an+1Xnx,f(Xn)+a2n+1f(Xn)2<Xnx22an+1Xnx,f(Xn)+an+1ϵˉf2. (2.26)

    6. We now consider two cases of (2.26). First, assume XnxR. Then (2.26) can immediately be bounded as

    Xn+1x2<R2+2ϵRˉf+ϵ2ˉf2=(R+ϵˉf)2<(R+1)2,

    where we have used condition (2.20) in the last inequality.

    7. Now consider the case R<Xnx<R+1. Using (2.20), the inner product in (2.26) can first be bounded from below:

    Xnx,f(Xn)=Xnx,f(Xn)+XnXn,f(Xn)Xnx,f(Xn)ϵˉf>Xnx,f(Xn)12δ.

    Next, using (2.20)

    XnxXnxXnXn>Rϵ>R12R=12R

    Therefore, 12R<Xnx<R+2, so (2.19) ensures Xnx,f(Xn)δ and

    Xnx,f(Xn)>δ12δ=12δ.

    Returning to (2.26), by (2.20),

    Xn+1x2(R+1)2an+1(δϵˉf2)<(R+1)2.

    This completes the proof of the inductive step in this second case, completing the proof.

    Corollary 2.1. For the expanding trust region algorithm, if Assumptions 1, 2, 3, and 4 hold, then the number of truncations is a.s. finite.

    Proof. The proof is by contradiction. We break the proof into 4 steps:

    1. Assuming that there are infinitely many truncations, Lemma 3 implies that the sequence cannot remain in a bounded set. Then, continuing to assume that Assumptions 1, 2, 3, and 4 hold, the only way for the conclusion of Lemma 4 to fail is if the assumption on Pn+11Xnxr is false. Therefore, there exists r>0 and a set of positive measure on which a subsequence, Pnk+11Xnkxr0. Hence XnkBr(x), and Pnk+10. So truncations occur at these indices, and ˜Xnk+1Uσnk.

    2. Let ˉf=supf(x) with the supremum over the set Br(x) and let ϵ>0 satisfy

    ϵ<(ˉf+1)1. (2.27)

    By our assumptions of the lemma and Proposition 1, there exists nϵ such that for all nnϵ

    an+1δMn+11Xnxrϵ,an+1ϵ (2.28)

    Along the subsequence, for all nknϵ,

    ank+1δMnk+11Xnkxr=ank+1δMnk+1ϵ. (2.29)

    3. Furthermore, for nknϵ:

    ˜Xnk+1xXnkx+ank+1f(Xnk)+ank+1δMnk+1<r+ϵˉf+ϵ<r+1,˜Xnk+1Br+1(x), (2.30)

    where (2.27) has been used in the last inequality.

    4. By the definition of the Un, there exists an index M such that UMBr+1(x). Let

    N=inf{nnϵσnM}. (2.31)

    This set is nonempty and N< since we have assumed there are infinitely many truncations. Let nkN. Then σnkM and UσnkBr+1(x). But (2.30) then implies that ˜Xnk+1Uσnk, and no truncation will occur; Pnk+1=0, providing the desired the contradiction.

    Using the above results, we are able to prove Theorems 2.1 and 2.2. Since the proofs are quite similar, we present the more complicated expanding trust region case.

    Proof. We split this proof into 6 steps:

    1. First, by Corollary 1, only finitely many truncations occur. By Lemma 3, there exists R>0 such that XnBR(0) for all n. Consequently, there is an r such that XnBr(x) for all n.

    2. Next, we fix constants. Let ˉf=supf(x) with the supremum taken over Br(x). Fix η(0,2R), and use Assumption 2 to determine δ>0 such that

    infη/2xxrxx,f(x)=δ (2.32)

    Take ϵ>0 such that:

    ϵ<min{1,η2,δ2ˉf,δ2ˉf2} (2.33)

    Having set ϵ, we again appeal to Assumption 4 and Proposition 1 to find nϵ such that for all nnϵ:

    i=n+1aiδMi1Xi1xr=i=n+1aiδMiϵ,an+1ϵ (2.34)

    3. Define the auxiliary sequence,

    Xn=Xni=n+1aiδMi1Xi1xr=Xni=n+1aiδMi. (2.35)

    Since there are only finitely many truncations, there exists Nnϵ, such that for all nN, Pn+1=0, as the truncations have ceased. Consequently, for nN,

    Xn+1=Xnan+1f(Xn) (2.36)

    By (2.34) and (2.35), for nN, XnXnϵ. Since ϵ>0 may be arbitrarily small, it will be sufficient to prove Xnx.

    4. To obtain convergence of Xn, we first examine Xn+1x. For nN,

    Xn+1x2Xnx22an+1Xnx,f(Xn)+an+1ϵˉf2, (2.37)

    Now consider two cases of this expression. First, assume Xnxη. In this case, using (2.33),

    2an+1Xnx,f(Xn)+an+1ϵˉf2an+1(2ηˉf+ϵˉf2)<an+1(4Rˉf+ˉf2)=an+1B. (2.38)

    where B>0 is a constant depending only on R and ˉf. For Xnx>η, using (2.33)

    Xnx,f(Xn)=Xnx,f(Xn)+XnXn,f(Xn)Xnx,f(Xn)ϵˉf>Xnx,f(Xn)12δ. (2.39)

    By (2.33),

    XnxXnxXnXn>ηϵ>12η

    Since Xnx<r too, (2.32) and (2.39) yield the estimate

    Xnx,f(Xn)>δϵˉf>12δ

    Thus, in this regime, using (2.33),

    2an+1Xnx,f(Xn)+an+1ϵˉf2an+1(δϵˉf2)<12δan+1=Aan+1 (2.40)

    where A>0 is a constant depending only on δ.

    Combining estimates (2.38) and (2.40), we can write for nN

    Xn+1x2<Xnx2an+1A1Xnx>η+an+1B1Xnxη. (2.41)

    5. We now show that Xnxη i.o. The argument is by contradiction. Let MN be such that for all nM, Xnx>η. For such n,

    η2<Xn+1x2<Xnx2an+1A<Xn1x2an+1AanA<<XMx2Ani=Mai+1. (2.42)

    Using Assumption 4 and taking n, we obtain a contradiction.

    6. Finally, we prove convergence of Xnx. Since XnBη(x) i.o., let

    N=inf{nNXnx<η}. (2.43)

    For nN, we can then define

    φ(n)=max{pnXpx<η}. (2.44)

    For all such n, φ(n)n, and Xφ(n)Bη(x).

    We claim that for nN,

    Xn+1x2<Xφ(n)x2+Baφ(n)+1<η2+Baφ(n)+1.

    First, if n=φ(n), this trivially holds in (2.41). Suppose now that n>φ(n). Then for i=φ(n)+1,φ(n)+2,n, Xix>η. Consequently,

    Xn+1x2<Xnx2<Xn1x2<<Xφ(n)+1x2<Xφ(n)x2+Baφ(n)+1<η2+Baφ(n)+1

    As φ(n),

    lim supnXn+1x2η2

    Since η may be arbitrarily small, we conclude that

    lim supnXn+1x=limnXn+1x=0,

    completing the proof.

    Recall from the introduction that our distribution of interest, μ, is posed on the Borel subsets of Hilbert space H. We assume that μμ0, where μ0=N(m0,C0) is some reference Gaussian. Thus, we write

    dμdμ0=1Zμexp{Φμ(u)}, (3.1)

    where Φν:XR, X a Banach space, a subspace of H, of full measure with respect to μ0, a Gaussian on H, assumed to be continuous. Zμ=Eμ0[exp{Φ(u)}](0,) is the partition function ensuring we have a probability measure.

    Let ν=N(m,C), be another Gaussian, equivalent to μ0, such that we can write

    dνdμ0=1Zνexp{Φν(v)}, (3.2)

    Assuming that νμ, we can write

    R(ν||μ)=Eν[Φμ(u)Φν(u)]+log(Zμ)log(Zν) (3.3)

    The assumption that \nu \ll \mu implies that \nu and \mu are equivalent measures. As was proven in [16], if \mathcal{A} is a set of Gaussian measures, closed under weak convergence, such that at least one element of \mathcal{A} is absolutely continuous with respect to \mu , then any minimizing sequence over \mathcal{A} will have a weak subsequential limit.

    If we assume, for this work, that C = C_0 , then, by the Cameron-Martin formula (see [9]),

    \begin{equation} \Phi_\nu(u) = -\left\langle {u-m},{m -m_0}\right\rangle_{ \mathcal{H}^1} - \frac{1}{2}\left \|{m - m_0}\right\|_{ \mathcal{H}^1}^2, \quad Z_{\nu} = 1. \end{equation} (3.4)

    Here, \left\langle {\bullet}, {\bullet}\right\rangle_{ \mathcal{H}^1} and \|\bullet\|_{ \mathcal{H}^1} are the inner product and norms of the Cameron-Martin Hilbert space, denoted \mathcal{H}^1 ,

    \begin{equation} \left\langle {f},{g}\right\rangle_{ \mathcal{H}^1} = \left\langle {C_0^{-1/2} f},{C_0^{-1/2} g}\right\rangle, \quad \left \|{f}\right\|_{ \mathcal{H}^1}^2 = \left\langle {f},{f}\right\rangle_{ \mathcal{H}^1}^2. \end{equation} (3.5)

    Convergence to the minimizer will be established in \mathcal{H}^1 , and \mathcal{H}^1 will be the relevant Hilbert space in our application of Theorems 2.1 and 2.2 to this problem.

    Letting \nu_0 = N(0, C_0) and v\sim \nu_0 , we can then rewrite (3.3) as

    \begin{equation} J(m) \equiv \mathcal{R}(\nu||\mu) = \mathbb{E}^{\nu_0}[\Phi_{\mu}(v + m)] + \frac{1}{2}\left \|{m - m_0}\right\|_{ \mathcal{H}^1}^2 + \log(Z_\mu) \end{equation} (3.6)

    The Euler-Lagrange equation associated with (3.6), and the second variation, are:

    \begin{align} J'(m) & = \mathbb{E}^{\nu_0}[\Phi'_\mu(v+m)] + C_0^{-1}(m-m_0), \end{align} (3.7)
    \begin{align} J''(m) & = \mathbb{E}^{\nu_0}[\Phi''_\mu(v+m)] + C_0^{-1}. \end{align} (3.8)

    In [15], it was suggested that rather than try to find a root of (3.7), the equation first be preconditioned by multiplying by C_0 ,

    \begin{equation} C_0 \mathbb{E}^{\nu_0}[\Phi'_\mu(v+m)] + (m-m_0), \end{equation} (3.9)

    and a root of this mapping is sought, instead. Defining

    \begin{align} f(m) & = C_0 \mathbb{E}^{\nu_0}[\Phi'_\mu(v+m)] + (m-m_0), \end{align} (3.10)
    \begin{align} F(m,v) & = C_0\Phi'_\mu(v+m) + (m-m_0). \end{align} (3.11)

    The Robbins-Monro formulation is then

    \begin{equation} m_{n+1} = m_n - a_{n+1} F(m_n, v_{n+1}) + P_{n+1}, \end{equation} (3.12)

    with v_n \sim \nu_0 , i.i.d.

    We thus have

    Theorem 3.1. Assume:

    There exists \nu = N(m, C_0)\sim \mu_0 such that \nu\ll\mu .

    \Phi_\mu' and \Phi_\mu'' exist for all u \in \mathcal{H}^1 .

    There exists m_\star , a local minimizer of J , such that J'(m_\star) = 0 .

    The mapping

    \begin{equation} m\mapsto \mathbb{E}^{\nu_0}\left[{\left \|{\sqrt{C_0}\Phi_\mu'(m+v)}\right\|^2}\right] \end{equation} (3.13)

    is bounded on bounded subsets of \mathcal{H}^1 .

    There exists a convex neighborhood U_\star of m_\star and a constant \alpha > 0 , such that for all m\in U_\star , for all u \in \mathcal{H}^1 ,

    \begin{equation} \left\langle {J''(m)u},{u}\right\rangle\geq \alpha \left \|{u}\right\|_{ \mathcal{H}^1}^2 \end{equation} (3.14)

    Then, choosing a_n according to Assumption 4,

    If the subset U_\star can be taken to be all of \mathcal{H}^1 , for the expanding truncation algorithm, m_n \to m_\star a.s. in \mathcal{H}^1 .

    If the subset U_\star is not all of \mathcal{H}^1 , then, taking U_1 to be a bounded (in \mathcal{H}^1 ) convex subset of U_\star , with m_\star \in U_1 , and U_0 any subset of U_1 such that there exist R_0 < R_1 with

    U_0 \subset B_{R_0}(x_\star) \subset B_{R_1}(x_\star)\subset U_1,

    for the fixed truncation algorithm, m_n\to m_\star a.s. in \mathcal{H}^1 .

    Proof. We split the proof into 2 steps:

    1. By the assumptions of the theorem, we clearly satisfy Assumptions 1 and 4. To satisfy Assumption 3, we observe that

    \begin{equation*} \mathbb{E}^{\nu_0}[\left \|{F(m,v)}\right\|^2_{ \mathcal{H}^1}]\leq 2 \mathbb{E}^{\nu_0}\left[{\left \|{\sqrt{C_0}\Phi_\mu'(m+v)}\right\|^2}\right] + 2\left \|{m-m_0}\right\|_{ \mathcal{H}^1}^2. \end{equation*}

    This is bounded on bounded subsets of \mathcal{H}^1 .

    2. Per the convexity assumption, (3.14), implies Assumption 2, since, by the mean value theorem in function spaces,

    \begin{equation*} \begin{split} \left\langle {m-m_\star},{f(m)}\right\rangle_{ \mathcal{H}^1} & = \left\langle {m-m_\star},{C_0\left[{J'(m_\star) +J''(\tilde m)(m-m_\star) }\right]}\right\rangle_{ \mathcal{H}^1}\\ & = \left\langle {m-m_\star},{J''(\tilde m)(m-m_\star)}\right\rangle\geq \alpha\left \|{m-m_\star}\right\|_{ \mathcal{H}^1}^2 \end{split} \end{equation*}

    where \tilde m is some intermediate point between m and m_\star . This completes the proof.

    While condition (3.14) is sufficient to obtain convexity, other conditions are possible. For instance, suppose there is a convex open set U_\star containing m_\star and constant \theta\in [0, 1) , such that for all m \in U_\star ,

    \begin{equation} \inf\limits_{\substack{u\in \mathcal{H}\\ u\neq 0}} \frac{\left\langle { \mathbb{E}^{\nu_0}[\Phi''_\mu(v+m)]u},{u}\right\rangle}{\left \|{u}\right\|^2}\geq -\theta\lambda_1^{-1}, \end{equation} (3.15)

    where \lambda_1 is the principal eigenvalue of C_0 . Then this would also imply Assumption 2, since

    \begin{equation*} \begin{split} \left\langle {m-m_\star},{f(m)}\right\rangle_{ \mathcal{H}^1} & = \left\langle {m-m_\star},{C_0\left[{J'(m_\star) +J''(\tilde m)(m-m_\star) }\right]}\right\rangle_{ \mathcal{H}^1}\\ & = \left\langle {m-m_\star},{J''(\tilde m)(m-m_\star)}\right\rangle\\ &\geq \left \|{m-m_\star}\right\|_{ \mathcal{H}^1}^2 + \left\langle {m-m_\star},{ \mathbb{E}^{\nu_0}[\Phi''_\mu(v+\tilde m)] (m-m_\star)}\right\rangle\\ &\geq \left \|{m-m_\star}\right\|_{ \mathcal{H}^1}^2 -\theta \lambda_1^{-1} \left \|{m-m_\star}\right\|^2\\ &\geq (1-\theta)\left \|{m-m_\star}\right\|_{ \mathcal{H}^1}^2. \end{split} \end{equation*}

    We mention (3.15) as there may be cases, shown below, for which the operator \mathbb{E}^{\nu_0}[\Phi''_\mu(v+ m)] is obviously nonnegative.

    To apply the Robbins-Monro algorithm to the relative entropy minimization problem, the \Phi_\mu functional of interest must be examined. In this section we present a few examples, based on those presented in [15], and examine when the assumptions hold. The one outstanding assumption that we must make is that, a priori, \mu_0 is an equivalent measure to \mu .

    Taking \mu_0 = N(0, 1) , the standard unit Gaussian, let V: \mathbb{R} \to \mathbb{R} be a smooth function such that

    \begin{equation} \frac{d\mu}{d\mu_0} = \frac{1}{Z_\mu} \exp\left\{{-{ \epsilon^{-1}}V(x)}\right\} \end{equation} (4.1)

    is a probability measure on \mathbb{R} . For these scalar cases, we use x in place of v . In the above framework,

    \begin{align*} F(x,\xi) & = { \epsilon^{-1}}V'(x+\xi) -\xi,\\ f(x) & = { \epsilon^{-1}} \mathbb{E}[V'(x+\xi)]m \\ \Phi_\mu'(x) & = { \epsilon^{-1}}V'(x), \\ \Phi_\mu''(x)& = { \epsilon^{-1}}V''(x) \end{align*}

    and \xi\sim N(0, 1) = \nu_0 = \mu_0 .

    Consider the case that

    \begin{equation} V(x) = \tfrac{1}{2}x^2 + \tfrac{1}{4}x^4. \end{equation} (4.2)

    In this case

    \begin{align*} F(x,\xi)& = { \epsilon^{-1}}\left({x+\xi + (x+\xi)^3}\right) + x,\\ f(x) & = { \epsilon^{-1}}\left({4x + x^3}\right) +x, \\ \mathbb{E}[\Phi''_\mu(x+\xi)] & = { \epsilon^{-1}}(4 + 3x^2),\\ \mathbb{E}[\left |{\Phi'_\mu(x+\xi)}\right |^2] & = { \epsilon^{-1}} \left({22 + 58 x^2 + 17 x^4 + x^6}\right). \end{align*}

    Since \mathbb{E}[\Phi''_\mu(x+\xi)] \geq 4 { \epsilon^{-1}} , all of our assumptions are satisfied and the expanding truncation algorithm will converge to the unique root at x_\star = 0 a.s. See Figure 1 for an example of the convergence at \epsilon = 0.1 , U_{n} = (-n -1, n+1) , and always restarting at 0.5 .

    Figure 1.  Robbins-Monro applied to a globally convex scalar problem associated with (4.2) with \epsilon = 0.1 and expanding trust regions U_{n} = (-1-n, 1+n) .

    We refer to this as a "globally convex'' problem since \mathcal{R} is globally convex about the minimizer.

    In contrast to the above problem, some mimizers are only "locally'' convex. Consider the case the double well potential

    \begin{equation} V(x) = \tfrac{1}{4}(4-x^2)^2 \end{equation} (4.3)

    Now, the expressions for RM are

    \begin{align*} F(x,\xi) & = { \epsilon^{-1}}\left({(x+\xi)^3-4(x+\xi))}\right) + x,\\ f(x) & = { \epsilon^{-1}}\left({x^3-x}\right) +x, \\ \mathbb{E}[\Phi''_\mu(x+\xi)] & = { \epsilon^{-1}}\left({3x^2-1}\right),\\ \mathbb{E}[\left |{\Phi'_\mu(x+\xi)}\right |^2] & = { \epsilon^{-1}} (1 + x^2) (7 + 6 x^2 +x^4). \end{align*}

    In this case, f(x) vanishes at 0 and \pm \sqrt{1- \epsilon} , and J'' changes sign from positive to negative when x enters ({-\sqrt{(1- \epsilon)/3}, \sqrt{({1- \epsilon})/{3}}}) . We must therefore restrict to a fixed trust region if we want to ensure convergence to either of \pm\sqrt{1- \epsilon} .

    We ran the problem at \epsilon = 0.1 in two cases. In the first case, U_1 = (0.6, 3.0) and the process always restarts at 2 . This guarantees convergence since the second variation will be strictly postive. In the second case, U_1 = (-0.5, 1.5) , and the process always restarts at -0.1 . Now, the second variation can change sign. The results of these two experiments appear in Figure 2. For some random number sequences the algorithm still converged to \sqrt{1- \epsilon} , even with the poor choice of trust region.

    Figure 2.  Robbins-Monro applied to the nonconvex scalar problem associated with (4.3). Figure (a) shows the result with a well chosen trust region, while (b) shows the outcome of a poorly chosen trust region.

    Take \mu_0 = N(m_0(t), C_0) , with

    \begin{equation} C_0 = \left({-\frac{d^2}{dt^2}}\right)^{-1}, \end{equation} (4.4)

    equipped with Dirichlet boundary conditions on \mathcal{H} = L^2(0, 1) .* In this case the Cameron-Martin space \mathcal{H}^1 = H^1_0(0, 1) , the standard Sobolev space equipped with the Dirichlet norm. Let us assume m_0 \in H^1(0, 1) , taking values in \mathbb{R}^d .

    * This is the covariance of the standard unit Brownian bridge, Y_t = B_t - t B_1 .

    Consider the path space distribution on L^2(0, 1) , induced by

    \begin{equation} \frac{d\mu}{d\mu_0} = - \frac{1}{Z_\mu}\exp\left\{{-\Phi_\mu(v)}\right\}, \quad \Phi_\mu(u) = { \epsilon^{-1}}\int_0^1 V(v(t))dt, \end{equation} (4.5)

    where V: \mathbb{R}^d\to \mathbb{R} is a smooth function. We assume that V is such that this probability distribution exists and that \mu \sim \mu_0 , our reference measure.

    We thus seek an \mathbb{R}^d valued function m(t) \in H^1(0, 1) for our Gaussian approximation of \mu , satisfying the boundary conditions

    \begin{equation} m(0) = m_-,\quad m(1) = m_+. \end{equation} (4.6)

    For simplicity, take m_0 = (1-t)m_- + t m_+ , the linear interpolant between m_\pm . As above, we work in the shifted coordinated x(t) = m(t) - m_0(t)\in H^1_0(0, 1) .

    Given a path v(t)\in H^1_0 , by the Sobolev embedding, v is continuous with its L^\infty norm controlled by its H^1 norm. Also recall that for \xi \sim N(0, C_0) , in the case of \xi(t) \in \mathbb{R} ,

    \begin{equation} \mathbb{E}\left[{\xi(t)^p}\right] = \begin{cases} 0, & \text{$p$ odd},\\ (p-1)!!\left[{t(1-t)}\right]^{\frac p 2}, & \text{$p$ even}. \end{cases} \end{equation} (4.7)

    Letting \lambda_1 = 1/\pi^2 be the ground state eigenvalue of C_0 ,

    \begin{equation*} \begin{split} \mathbb{E}[\|\sqrt{C_0}\Phi'_\mu(v +m_0+\xi)\|^2]&\leq {\lambda_1} \mathbb{E}[\|\Phi'_\mu(v +m_0+\xi)\|^2]\\ &\quad = {\lambda_1}{ \epsilon^{-2}}\int_0^1 \mathbb{E}[{\left |{V'(v(t)+m_0(t)+\xi(t))}\right |^2}]dt. \end{split} \end{equation*}

    The terms involving v+m_0 in the integrand can be controlled by the L^\infty norm, which in turn is controlled by the H^1 norm, while the terms involving \xi can be integrated according to (4.7). As a mapping applied to x , this expression is bounded on bounded subsets of H^1 .

    Minimizers will satisfy the ODE

    \begin{equation} { \epsilon}^{-1} \mathbb{E}\left[{V'(x+m_0 +\xi)}\right] -x'' = 0,\quad x(0) = x(1) = 0. \end{equation} (4.8)

    With regard to convexity about a minimizer, m_\star , if, for instance, V'' were pointwise positive definite, then the problem would satisfy (3.15), ensuring convergence. Consider the quartic potential V given by (4.2). In this case,

    \begin{equation} \Phi(v) = { \epsilon}^{-1}\int_0^1 \frac{1}{2}v(t)^2 +\frac{1}{4}v(t)^4 dt, \end{equation} (4.9)

    and

    \begin{align*} \Phi'(v+m_0+ \xi) & = { \epsilon}^{-1}\left[{(v+m_0 +\xi) +3(v+m_0 +\xi)^3 }\right]\\ \Phi''(v+m_0 + \xi) & = { \epsilon}^{-1}\left[{1 +3 (v+m_0+\xi)^2}\right],\\ \mathbb{E}[\Phi'(v+m_0 + \xi)]& = { \epsilon}^{-1}\left[{v+m_0 +(v+m_0)^3+ 3 t(1-t) (v+m_0)}\right]\\ \mathbb{E}[\Phi''(v+m_0 + \xi)]& = { \epsilon}^{-1}\left[{1 + 3 (v+m_0)^2 + 3 t(1-t) }\right] \end{align*}

    Since \Phi''(v+m_0+\xi)\geq \epsilon^{-1} , we are guaranteed convergence using expanding trust regions. Taking \epsilon = 0.01 , m_- = 0 and m_+ = 2 , this is illustrated in Figure 3, where we have also solved (4.8) by ODE methods for comparison. As trust regions, we take

    \begin{equation} U_n = \left\{{m \in H^1_0(0,1)\mid \left \|{x}\right\|_{H^1}\leq 10+n}\right\}, \end{equation} (4.10)
    Figure 3.  The mean paths computed for (4.9) at different resolutions, along with the truncation sequence.

    and we always restart at the zero solution Figure 3 also shows robustness to discretization; the number of truncations is relatively insensitive to \Delta t .

    For many problems of interest, we do not have global convexity. Consider the double well potential (4.3), but in the case of paths,

    \begin{equation} \Phi(u) = { \epsilon^{-1}}\int_0^1\frac{1}{4} (4-v(t)^2)^2dt. \end{equation} (4.11)

    Then,

    \begin{align*} \Phi'(v + m_0 + \xi)& = { \epsilon}^{-1}\left[{(v + m_0 + \xi)^3 - 4 (v + m_0 + \xi)}\right]\\ \Phi''(v + m_0 + \xi) & = { \epsilon}^{-1}\left[{3 (v + m_0 + \xi)^2 - 4}\right],\\ \mathbb{E}[\Phi'(v+m_0 + \xi)]& = { \epsilon}^{-1}\left[{(v+m_0)^3 + 3 t(1-t) (v+m_0)-4(v+m_0)}\right]\\ \mathbb{E}[\Phi''(v+m_0 + \xi)]& = { \epsilon}^{-1}\left[{3(v+m_0)^2 + 3 t(1-t) -4}\right] \end{align*}

    Here, we take m_- = 0 , m_+ = 2 , and \epsilon = 0.01 . We have plotted the numerically solved ODE in Figure 4. Also plotted is \mathbb{E}[\Phi''(v_\star +m_0+ \xi)] . Note that \mathbb{E}[\Phi''(v_\star +m_0+ \xi)] is not sign definite, becoming as small as -400 . Since C_0 has \lambda_1 = 1/\pi^2 \approx 0.101 , (3.15) cannot apply.

    Figure 4.  The numerically computed solution to (4.8) in the case of the double well, (4.11), m_\star , and the associated \mathbb{E}^{\nu_0}[\Phi''(m_\star + \xi)] .

    Discretizing the Schrödinger operator

    \begin{equation} J''(v_\star) = -\frac{d^2}{dt^2} + { \epsilon}^{-1}\left({3(v_\star(t)+m_0(t))^2 + 3 t(1-t) -4}\right), \end{equation} (4.12)

    we numerically compute the eigenvalues. Plotted in Figure 5, we see that the minimal eigenvalue of J''(m_\star) is approximately \mu_1\approx 550 . Therefore,

    \begin{equation} \left\langle {J''(x_\star)u},{u}\right\rangle\geq \mu_1 \left \|{u}\right\|^2_{L^2}\Rightarrow \left\langle {J''(x)u},{u}\right\rangle\geq \alpha\left \|{u}\right\|_{H^1}^2, \end{equation} (4.13)
    Figure 5.  The numerically computed spectrum for (4.12), associated with the m_\star shown in Figure 4. Also shown is the numerically computed spectrum for the path m(t) = 2t^2 , which introduces negative eigenvalues.

    for all v in some neighborhood of v_\star . For an appropriately selected fixed trust region, the algorithm will converge.

    However, we can show that the convexity condition is not global. Consider the path m(t) = 2t^2 , which satisfies the boundary conditions. As shown in Figure 5, this path induces negative eigenvalues.

    Despite this, we are still observe convergence. Using the fixed trust region

    \begin{equation} U_1 = \left\{{x\in H^1_0(0,1)\mid \left \|{x}\right\|_{H^1}\leq 100}\right\}, \end{equation} (4.14)

    we obtain the results in Figure 6. Again, the convergence is robust to discretization.

    Figure 6.  The mean paths computed for (4.11) at different resolutions, along with the truncation sequence.

    We have shown that the Robbins-Monro algorithm, with both fixed and expanding trust regions, can be applied to Hilbert space valued problems, adapting the finite dimensional proof of [12]. We have also constructed sufficient conditions for which the relative entropy minimization problem fits within this framework.

    One problem we did not address here was how to identify fixed trust regions. Indeed, that requires a tremendous amount of a priori information that is almost certainly not available. We interpret that result as a local convergence result that gives a theoretical basis for applying the algorithm. In practice, since the root is likely unknown, one might run some numerical experiments to identify a reasonable trust region, or just use expanding trust regions. The practitioner will find that the algorithm converges to a solution, though perhaps not the one originally envisioned. A more sophisticated analysis may address the convergence to a set of roots, while being agnostic as to which zero is found.

    Another problem we did not address was how to optimize not just the mean, but also the covariance in the Gaussian. As discussed in [15], it is necessary to parameterize the covariance in some way, which will be application specific. Thus, while the form of the first variation of relative entropy with respect to the mean, (3.7), is quite generic, the corresponding expression for the covariance will be specific to the covariance parameterization. Additional constraints are also necessary to guarantee that the parameters always induce a covariance operator. We leave such specialization as future work.

    This work was supported by US Department of Energy Award DE-SC0012733. This work was completed under US National Science Foundation Grant DMS-1818716. The authors would like to thank J. Lelong for helpful comments, along with anonymous reviewers whose reports significantly impacted our work.

    The authors declare that there is no conflicts of interest in this paper.



    [1] Van Huylenbroeck G, Vandermeulen V, Mettepenningen E, et al. (2007) Multifunctionality of agriculture: A review of definitions, evidence and instruments. Living Rev Landsc Res 1: 1–43. https://doi.org/10.12942/lrlr-2007-3 doi: 10.12942/lrlr-2007-3
    [2] Rogers E M (1983) Diffusion of innovations. 3rd Ed., New York: Free Press, 15.
    [3] Hermans F, Klerkx L, Roep D (2015) Structural conditions for collaboration and learning in innovation networks: Using an innovation system performance lens to analyze agricultural knowledge systems. J Agric Educ Ext 21: 35–54. https://doi.org/10.1080/1389224X.2014.991113 doi: 10.1080/1389224X.2014.991113
    [4] Lioutas ED, Charatsari C, Černič Istenič M, et al. (2019) The challenges of setting up the evaluation of extension systems by using a systems approach: the case of Greece, Italy and Slovenia. J Agric Educ Ext 25: 139–160. https://doi.org/10.1080/1389224X.2019.1583818 doi: 10.1080/1389224X.2019.1583818
    [5] Klerkx L, Van Mierlo B, Leeuwis C (2012) Evolution of systems approaches to agricultural innovation: Concepts, analysis, and interventions, In: Darnhofer I, Gibbon D, Dedieu B (Eds.), Farming Systems Research into the 21st Century: The New Dynamic, Springer Dordrecht, 457–483. https://doi.org/10.1007/978-94-007-4503-2_20
    [6] Klerkx L, Leeuwis C (2008) Balancing multiple interests: Embedding innovation intermediation in the agricultural knowledge infrastructure. Technovation 28: 364–378. https://doi.org/10.1016/j.technovation.2007.05.005 doi: 10.1016/j.technovation.2007.05.005
    [7] EU SCAR (2012) Agricultural knowledge, and innovation systems in transition—a reflection paper, Brussels, p.13. Available from: https://ec.europa.eu/eip/agriculture/en/publications/agricultural-knowledge-and-innovation-systems.html.
    [8] EU SCAR AKIS (2019) Preparing for Future AKIS in Europe.
    [9] Klerkx L (2020) Advisory services and transformation, plurality and disruption of agriculture and food systems: Towards a new research agenda for agricultural education and extension studies. J Agric Educ Ext 26: 131–140. https://doi.org/10.1080/1389224X.2020.1738046 doi: 10.1080/1389224X.2020.1738046
    [10] Welter F (2011) Contextualizing entrepreneurship—conceptual challenges and ways forward. Entrep Theory Pract 35: 165–184. https://doi.org/10.1111/j.1540-6520.2010.00427.x doi: 10.1111/j.1540-6520.2010.00427.x
    [11] McElwee G, Smith R (2014) Chapter 14 Researching rural enterprise. In: Fayolle A (Ed.), Handbook of Research On Entrepreneurship: What We Know and What We Need to Know, Edward Elgar Publishing, 307. https://doi.org/10.4337/9780857936929.00022
    [12] Labarthe P, Laurent C (2013) Privatization of agricultural extension services in the EU: Towards a lack of adequate knowledge for small-scale farms? Food Policy 38: 240–252. https://doi.org/10.1016/j.foodpol.2012.10.005 doi: 10.1016/j.foodpol.2012.10.005
    [13] Davis K (2019) Agricultural Extension and Education for the Future, Closing keynote in 24th European Seminar on Extension and Education. Available from: https://www.google.com/url?sa = t & rct = j & q = & esrc = s & source = web & cd = & ved = 2ahUKEwiRjPnbrJuBAxW3xwIHHVGeA1gQFnoECBAQAQ & url = https%3A%2F%2Fwww.reterurale.it%2Fflex%2Fcm%2Fpages%2FServeAttachment.php%2FL%2FIT%2FD%2F2%25252Fb%25252F2%25252FD.b2f20be86583c28041cc%2FP%2FBLOB%253AID%253D19744%2FE%2Fpdf & usg = AOvVaw0tuNdggviVs-Zu1iIVxlLW & opi = 89978449.
    [14] EU Commission (2012) The future of food and farming.
    [15] ASHBY J (2009) Fostering farmer first methodological innovation: Organizational learning and change in international agricultural research. Farmer First Revisited: Innovation Agric Res Dev 2009: 39–45.
    [16] Sutherland LA, Madureira L, Dirimanova V, et al. (2017) New knowledge networks of small-scale farmers in Europe's periphery. Land Use Policy 63: 428–439. https://doi.org/10.1016/j.landusepol.2017.01.028 doi: 10.1016/j.landusepol.2017.01.028
    [17] EU Commission (2009) Report from the Commission to the European Parliament and the Council on the Application of the Farm Advisory System as Defined in Article 12 and 13 of Council Regulation (EC) No 73/2009. Available from: ttps: //eurlex.europa.eu/LexUriServ/LexUriServ.do?uri = OJ: L: 2009: 030: 0016: 0099: en: PDF.
    [18] Knierim A, Labarthe P, Laurent C, et al. (2017) Pluralism of agricultural advisory service providers—Facts and insights from Europe. J Rural Stud 55: 45–58. https://doi.org/10.1016/j.jrurstud.2017.07.018 doi: 10.1016/j.jrurstud.2017.07.018
    [19] Vecchio Y, De Castro P, Masi M, et al. (2021) Do rural development policies really help small farms? A reflection from Italy. EuroChoices 20: 75–80. https://doi.org/10.1111/1746-692X.12338 doi: 10.1111/1746-692X.12338
    [20] van der Ploeg JD, Barjolle D, Bruil J, et al. (2019) The economic potential of agroecology: Empirical evidence from Europe. J Rural Stud 71: 46–61. https://doi.org/10.1016/j.jrurstud.2019.09.003 doi: 10.1016/j.jrurstud.2019.09.003
    [21] Birner R, Davis K, Pender J, et al. (2009) From best practice to best fit: A framework for designing and analyzing pluralistic agricultural advisory services worldwide. J Agric Educ Ext 15: 341–355. https://doi.org/10.1080/13892240903309595 doi: 10.1080/13892240903309595
    [22] Klerkx L, Petter Stræ te E, Kvam G T, et al. (2017) Achieving best-fit configurations through advisory subsystems in AKIS: case studies of advisory service provisioning for diverse types of farmers in Norway. J Agric Educ Ext 23: 213–229. https://doi.org/10.1080/1389224X.2017.1320640 doi: 10.1080/1389224X.2017.1320640
    [23] Marsden T, Sonnino R (2012) Human health and wellbeing and the sustainability of urban-regional food systems. Curr Opin Environ Sustain 4: 427–430. https://doi.org/10.1016/j.cosust.2012.09.004 doi: 10.1016/j.cosust.2012.09.004
    [24] Landini F, Brites W, Mathot y Rebolé M I (2017) Towards a new paradigm for rural extensionists' in-service training. J Rural Stud 51: 158–167. https://doi.org/10.1016/j.jrurstud.2017.02.010 doi: 10.1016/j.jrurstud.2017.02.010
    [25] OECD (2000) Multifunctionality. Towards an Analytical Framework.
    [26] Wilson G (2008) From 'weak' to 'strong' multifunctionality: Conceptualizing farm-level multifunctional transitional pathways. J Rural Stud 24: 367–383. https://doi.org/10.1016/j.jrurstud.2007.12.010 doi: 10.1016/j.jrurstud.2007.12.010
    [27] Poppe K (2014) The role of the European innovation partnership in linking innovation and research in agricultural knowledge and innovation systems. Agriregionieuropa 10: 37.
    [28] Foti R, Nyakudya I, Moyo M, et al. (2007) Determinants of farmer demand for 'fee-for-service' extension in Zimbabwe: The case of Mashonaland Central province. J Agric Educ Ext 14: 95–104. https://doi.org/10.5191/jiaee.2007.14108 doi: 10.5191/jiaee.2007.14108
    [29] Vanni F (2014) Agriculture and Public Goods: The Role of Collective Action. https://doi.org/10.1007/978-94-007-7457-5
    [30] Acheson JM (2006) Institutional failure in resource management. Annu Rev Anthropol 35: 117–134. https://doi.org/10.1146/annurev.anthro.35.081705.123238 doi: 10.1146/annurev.anthro.35.081705.123238
    [31] Andrew B (2008) Market failure, government failure and externalities in climate change mitigation: The case for a carbon tax. Public Adm Dev 28: 393–401. https://doi.org/10.1002/pad.517 doi: 10.1002/pad.517
    [32] Stiglitz J (2009) Government failure vs. market failure: Principles of regulation. In: Balleisen E, Moss D (Eds.), Government and Markets: Toward a New Theory of Regulation, Cambridge: Cambridge University Press, 13–51. https://doi.org/10.1017/CBO9780511657504.002
    [33] Randall A (2002) Valuing the outputs of multifunctional agriculture. Eur Rev Agric Econ 29: 289–307. https://doi.org/10.1093/eurrag/29.3.289 doi: 10.1093/eurrag/29.3.289
    [34] Eastwood C, Klerkx L, Nettle R (2017) Dynamics and distribution of public and private research and extension roles for technological innovation and diffusion: Case studies of the implementation and adaptation of precision farming technologies. J Rural Stud 49: 1–12. https://doi.org/10.1016/j.jrurstud.2016.11.008 doi: 10.1016/j.jrurstud.2016.11.008
    [35] Faure G, Desjeux Y, Gasselin P (2012) New challenges in agricultural advisory services from a research perspective: A literature review, synthesis and research agenda. J Agric Educ Ext 18: 461–492. https://doi.org/10.1080/1389224X.2012.707063 doi: 10.1080/1389224X.2012.707063
    [36] Vargo SL, Lusch RF (2017) Service-dominant logic 2025. Int J Res Mark 34: 46–67. https://doi.org/10.1016/j.ijresmar.2016.11.001 doi: 10.1016/j.ijresmar.2016.11.001
    [37] Prager K, Labarthe P, Caggiano M, et al. (2016) How does commercialization impact on the provision of farm advisory services? Evidence from Belgium, Italy, Ireland and the UK. Land Use Policy 52: 329–344. https://doi.org/10.1016/j.landusepol.2015.12.024 doi: 10.1016/j.landusepol.2015.12.024
    [38] Ward JH (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58: 236–244. https://doi.org/10.1080/01621459.1963.10500845 doi: 10.1080/01621459.1963.10500845
    [39] Pigford AAE, Hickey GM, Klerkx L (2018) Beyond agricultural innovation systems? Exploring an agricultural innovation ecosystems approach for niche design and development in sustainability transitions. Agric Syst 164: 116–121. https://doi.org/10.1016/j.agsy.2018.04.007 doi: 10.1016/j.agsy.2018.04.007
    [40] Van der Ploeg JD, Marsden T (2008) Unfolding webs: The dynamics of regional rural development van Gorcum, Assen.
    [41] Renting H, Rossing WAH, Groot JCJ, et al. (2009) Exploring multifunctional agriculture. A review of conceptual approaches and prospects for an integrative transitional framework. J Environ Manage 90: S112–S123. https://doi.org/10.1016/j.jenvman.2008.11.014 doi: 10.1016/j.jenvman.2008.11.014
  • 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(2084) PDF downloads(141) Cited by(1)

Figures and Tables

Figures(1)  /  Tables(6)

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog