Abstract
This is the first paper dealing with the study of minimum and maximum principle sufficiency properties for nonsmooth variational inequalities by using gap functions in the setting of Hadamard manifolds. We also provide some characterizations of these two sufficiency properties. We conclude the paper with a discussion of the error bounds for nonsmooth variational inequalities in the setting of Hadamard manifolds.
Similar content being viewed by others
1 Introduction
The variational inequality problem, introduced by Hartman and Stampacchia [20], plays an important role in current mathematical technology. It has been extended and generalized to study a wide class of problems arising in optimization, nonlinear programming, physics, economics, transportation equilibrium problems, and engineering sciences. For further details, see [1–4, 7, 8, 21–23, 25, 32, 36] and the references therein.
The weak sharp minimum property for the convex optimization problem was introduced and studied by Ferris [13, 17]. It has significant applications in sensitivity analysis, error bounds, and finite convergence of algorithms for solving convex optimization problems, see [12, 17, 18, 27]. In 1992, Ferris and Mangasarian [19] studied the minimum principle sufficiency property for convex programming and proved that this property is equivalent to the weak sharp minimum property of convex programming. Marcotte and Zhu [24] extended the minimum principle sufficiency property for variational inequalities and characterized the weak sharp solutions for variational inequalities using pseudomonotone+ and continuous mapping on a compact polyhedral set. Extending the results of Marcotte and Zhu [24], Wu and Wu [35] established the maximum principle sufficiency property for variational inequalities and also characterized the weak sharp solutions of variational inequalities. Recently, Wu [34] characterized the minimum principle sufficiency property when the mapping in a variational inequality is constant and pseudomonotone+.
The variational inequality problem can be used to solve any differentiable optimization problem over a convex feasible region. However, in many practical problems, the function involved in the optimization problem need not be differentiable but has some kind of directional derivative. In this case, the relevant optimization problem can be considered by using a nonsmooth variational inequality with a bifunction, see [10]. In 2016, Alshahrani et al. [6] defined the minimum and maximum principle sufficiency properties for nonsmooth variational inequalities (NVI) by using a gap function that is similar to that of Wu and Wu [35] and provided some characterizations for these two properties. They also discussed error bounds for nonsmooth variational inequalities. For further related results, see [5, 11, 33] and the reference therein.
On the other hand, in the last two decades, several classical problems have been extended from a linear space setting to Riemannian/Hadamard manifolds setting (nonlinear space) because some nonconvex and constrained optimization problems in Euclidean space become convex and unconstrained ones in Riemannian/Hadamard manifolds, see [31]. In 2003, Németh [25] extended the notion of variational inequalities to Hadamard manifolds and related it to geodesic convex optimization problems. He also proved the existence and uniqueness results for variational inequalities on Hadamard manifolds. Then, Colao et al. [15] developed the equilibrium theory in Hadamard manifolds and also proved some existence results. In 2017, we [9] derived the generalized convexity of a real-valued function defined on a Riemannian manifold in terms of a bifunction h. We also defined the generalized monotonicity of the bifunction h. We also proved the relationship between the generalized convexity of a real-valued function and the generalized monotonicity of h. In 2020, we [8] extended the notion of the nonsmooth variational inequality problem (NVIP) and the Minty nonsmooth variational inequality problem (MNVIP) in the setting of Hadamard manifolds and proved some existence results for the nonsmooth variational inequality problem. We gave some relations among (NVIP), (MNVIP), and optimization problems in the setting of Hadamard manifolds. The gap functions for (NVIP) and (MNVIP) were also studied in the setting of Hadamard manifolds.
Motivated by the above results, in the present paper, we extend the work of Alshahrani et al. [6] in the setting of Hadamard manifolds. Hence, we defined the minimum and maximum principle sufficiency properties for nonsmooth variational inequalities by using a gap function in the Hadamard manifolds setting and provide several characterizations for these two properties. We also study the error bounds for nonsmooth variational inequalities in the setting of Hadamard manifolds.
2 Preliminaries
In the present section, we recall some basic definitions, properties, notions and results from manifolds, which will be needed throughout the paper. For details, see [14, 16, 26, 28–31].
Let M be a finite-dimensional differentiable manifold. For each \(x\in M\), let \(T_{x}M\) be the tangent space at the point x to M, which is a real vector space of the same dimension as M. The collection of all tangent spaces on M is called a tangent bundle and it is denoted by TM. A \(C^{\infty }\) mapping \(A:M\rightarrow TM\), which assigns a tangent vector \(A(x)\) at x for each \(x\in M\), is called a vector field on M. We denote by \(\langle \cdot, \cdot \rangle _{x}\) the scalar product on \(T_{x}M\) with the associated norm \(\| \cdot \|_{x}\), where the subscript x can be omitted if there is no confusion.
A scalar product \(\langle \cdot, \cdot \rangle _{x}\) is called the Riemannian metric on \(T_{x}M\). A \(C^{\infty }\) tensor field g of type \((0, 2)\) on M is called the Riemannian metric on M if for each \(x\in M\), the tensor \(g(x)\) is a Riemannian metric on \(T_{x}M\), and the pair \((M, g)\) is called the Riemannian manifold.
For any \(x, y \in M\), let \(\gamma:[0, 1]\rightarrow M\) be a piecewise-smooth curve joining x to y. The arc length of γ is defined by
where \(\dot{\gamma }(t)\) denotes the tangent vector at \(\gamma (t)\). For any \(x, y \in M\), the Riemannian distance from x to y is defined by
where the infimum is taken over all piecewise-smooth curves \(\gamma:[0, 1]\rightarrow M\) joining x to y.
Let ∇ be the Levi–Civita connection on M. A curve \(\gamma: [0, 1]\rightarrow M\) joining x to y is said to be a geodesic if
A geodesic \(\gamma:[0, 1]\rightarrow M\) joining x to y is said to be minimal if its arc length equals its Riemannian distance between x and y.
A Riemannian manifold M is said to be complete if for any \(x\in M\), all the geodesics emanating from x are defined for all \(t\in \mathbb{R}\).
A simply connected complete Riemannian manifold M of nonpositive sectional curvature is called a Hadamard manifold [16].
Let M be a Hadamard manifold. The exponential mapping [31] \(\exp _{x}:T_{x}M\rightarrow M\) at x is defined by \(\exp _{x}v=\gamma _{v}(1, x)\) for each \(v\in T_{x}M\), where \(\gamma (\cdot )= \gamma _{v}(\cdot, x)\) is the geodesic starting at x with the velocity v, that is, \(\gamma (0) =x\) and \(\dot{\gamma }(0)=v\). Moreover, \(\exp _{x}{tv}=\gamma _{v}(t, x)\) for each real number t.
Proposition 1
([16])
Let M be a Hadamard manifold and \(x\in M\). Then, \(\exp _{x}: T_{x}M\rightarrow M\) is a diffeomorphism, and for any two points \(x, y\in M\), there exists a unique minimal geodesic joining x to y
for all \(t\in [0, 1]\).
In particular, the exponential mapping and its inverse are continuous on a Hadamard manifold.
Let M be a Hadamard manifold. We denote by \(P_{\gamma (b), \gamma (a)}\) the parallel transport from \(T_{\gamma (a)}M\) to \(T_{\gamma (b)}M\) along the geodesic γ with respect to ∇ and it is defined by
where A is the unique vector field such that \(\nabla _{\dot{\gamma }(t)}v=0\) for all t and \(A(\gamma (a))=v\).
Note that
-
(i)
\(P_{\gamma (a), \gamma (a)}=\mathrm{Id}\), the identity transformation of \(T_{\gamma (a)}M\).
-
(ii)
\(P_{\gamma (a), \gamma (b)}\circ P_{\gamma (b), \gamma (c)}=P_{ \gamma (a), \gamma (c)}\).
-
(iii)
\(P_{\gamma (a), \gamma (b)}\circ P_{\gamma (b), \gamma (a)}=P_{ \gamma (a), \gamma (a)}\).
Definition 1
A subset K of a Hadamard manifold M is said to be geodesic convex if for any pair of distinct points \(x, y\in K\), the geodesic γ joining x to y belongs to K, that is, if for any \(\gamma:[0, 1]\rightarrow M\) such that \(\gamma (0)=x\) and \(\gamma (1)=y\), then \(\gamma (t)=\exp _{x} (t\exp _{x}^{-1}y )\in K\) for all \(t\in [0, 1]\).
3 Formulation of the problems
Throughout the paper, unless otherwise specified, we assume that M is a Hadamard manifold, K is a nonempty geodesic convex subset of M and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) is a bifunction.
Recently, we [8] considered the following nonsmooth variational inequality problems:
Nonsmooth variational inequality problem (in short, NVIP): Find \(\bar{x}\in K\) such that
Minty-type nonsmooth variational inequality problem (in short, MTNVIP): Find \(\bar{x}\in K\) such that
We denote by \(S^{*}\) and \(S_{*}\) the solution set of NVIP (1) and MTNVIP (2), respectively, and assume that they are nonempty.
Consider the following optimization problem (in short, OP):
where \(f:K\rightarrow \mathbb{R}\) is a real-valued function. Assume that the solution set \(\bar{S}=\{ x\in K: f(x)\leq f(y) \text{ for all } y\in K\}\) of OP (3) is nonempty.
Definition 2
([8])
A bifunction \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) is said to be geodesic upper sign continuous if for any \(x, y\in K\),
where \(w=\exp _{y}(t\exp _{y}^{-1}x)\) for all \(t\in (0, 1)\).
Definition 3
([8])
A bifunction \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) is said to be pseudomonotone if for all \(x, y\in K\),
Recently, we [8] established the following equivalence result between the solution set of NVIP (1) and MTNVIP (2) under the pseudomonotonicity and geodesic upper sign continuity assumptions of the bifunction h.
Lemma 1
([8])
Let \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be a pseudomonotone and geodesic upper sign continuous bifunction such that h is positively homogeneous in the second argument, that is, for all \(\alpha >0\) and \(v\in T_{x}M\), \(h(x; \alpha v)=\alpha h(x; v)\). Then, \(\bar{x}\in K\) is a solution of NVIP (1) if and only if it is a solution of MTNVIP (2).
We consider the following condition, which was first considered by Wu and Wu [35], to develop the weak sharpness of variational inequalities in Hilbert spaces and later it was considered by Alshahrani et al. [6].
Condition A
For all \(x, y\in K\),
Note that \(\mathbb{U}(x)\) is nonempty for all \(x\in S^{*}\) and also \(\mathbb{V}(y)\) is nonempty for all \(y\in K\), since \(S_{*}\) is assumed to be nonempty. If h is pseudomonotone, then any \(\exp _{x}^{-1}y\) from \(\mathbb{U}(x)\) also belongs to \(\mathbb{V}(y)\) for all \(x, y\in K\).
The following result shows the equivalence between \(S^{*}\) and \(S_{*}\) without pseudomonotonicity and geodesic upper sign continuity of h, but under the assumption of Condition A.
Proposition 2
Let \(\bar{x}\in K\) and assume that \(\mathbb{U}(\bar{x})=\mathbb{V}(y)\) for all \(y\in K\). Then,
Proof
Let \(\bar{x}\in S^{*}\). Then, \(h(\bar{x}; \exp _{\bar{x}}^{-1}y)\geq 0\) for all \(y\in K\), and therefore, \(\exp _{\bar{x}}^{-1}y\in \mathbb{U}(\bar{x})\) for all \(y\in K\). By hypothesis, \(\mathbb{U}(\bar{x})=\mathbb{V}(y)\) for all \(y\in K\) and thus, \(\exp _{\bar{x}}^{-1}y\in \mathbb{V}(y)\) for all \(y\in K\). Hence,
Since \(P_{y, \bar{x}}\exp _{\bar{x}}^{-1}y=-\exp _{y}^{-1}\bar{x}\), we conclude that \(h(y; \exp _{y}^{-1}\bar{x})\leq 0\) for all \(y\in K\). Hence, \(\bar{x}\in S_{*}\).
Conversely, assume that \(\bar{x}\in S_{*}\). Then, \(h(y; \exp _{y}^{-1}\bar{x})\leq 0\) for all \(y\in K\). Since \(P_{y, \bar{x}}\exp _{\bar{x}}^{-1}y=-\exp _{y}^{-1}\bar{x}\), we have
Therefore, \(\exp _{\bar{x}}^{-1}y\in \mathbb{V}(y)\) for all \(y\in K\). By hypothesis, \(\mathbb{U}(\bar{x})=\mathbb{V}(y)\) for all \(y\in K\), therefore we have \(\exp _{\bar{x}}^{-1}y\in \mathbb{U}(\bar{x})\), that is,
and hence, \(\bar{x}\in S^{*}\). □
Definition 4
([8])
A function \(g: K\rightarrow \mathbb{R}\) is said to be a gap function for NVIP (1) (respectively, MTNVIP (2)) if it satisfies the following properties:
-
(a)
\(g(x)\geq 0\) for all \(x\in K\);
-
(b)
\(g(\bar{x})=0\) if and only if \(\bar{x}\in K\) is a solution of NVIP (1) (respectively, MTNVIP (2)).
The primal gap function \(\varphi (x)\) associated with NVIP (1) is defined by
and we set
In a similar way, the dual gap function \(\Phi (x)\) associated with MTNVIP (2) is defined by
and we set
Note that both the functions φ and Φ are nonnegative on K and vanish on \(S^{*}\) and \(S_{*}\), respectively. Therefore, they are also gap functions for NVIP (1) and MTNVIP (2), respectively.
4 Characterizations of solution sets
Definition 5
([9])
A function \(f:K\rightarrow \mathbb{R}\) is said to be geodesic radially upper semicontinuous (respectively, geodesic radially lower semicontinuous) on K if for every pair of distinct points \(x, y\in K\), the function f is upper semicontinuous (respectively, lower semicontinuous) along the geodesic segment \(\gamma _{xy}(t)\) for all \(t\in [0, 1]\), that is, \(t\mapsto f(\gamma _{xy}(t))\) is upper semicontinuous (respectively, lower semicontinuous) on \([0, 1]\). The function f is said to be geodesic radially continuous on K if it is both geodesic radially upper semicontinuous as well as geodesic radially lower semicontinuous on K.
Definition 6
([9])
Let \(f:M\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be a function on a Hadamard manifold M and x be a point where f is finite.
-
(a)
The Dini-upper directional derivative at \(x\in M\) in the direction \(v\in T_{x}M\) is defined by
$$\begin{aligned} f^{D}(x; v)=\limsup_{t\to 0^{+}}\frac{f(\exp _{x}tv)-f(x)}{t}. \end{aligned}$$ -
(b)
The Dini-lower directional derivative at \(x\in M\) in the direction \(v\in T_{x}M\) is defined by
$$\begin{aligned} f_{D}(x; v)=\liminf_{t\to 0^{+}}\frac{f(\exp _{x}tv)-f(x)}{t}. \end{aligned}$$
Definition 7
([9])
A function \(f:K\rightarrow \mathbb{R}\) is said to be geodesic h-pseudoconvex if for any pair of distinct points \(x, y\in K\), we have
equivalently,
The function f is called geodesic h-pseudoconcave if in Definition (7), the inequality < is replaced by >.
Definition 8
([9])
A function \(f:K\rightarrow \mathbb{R}\) is said to be geodesic h-pseudolinear if f is both geodesic h-pseudoconvex as well as geodesic h-pseudoconcave.
Theorem 1
([9])
Let \(f:K\rightarrow \mathbb{R}\) be a function and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous in the second argument such that
Further, assume that one of the following conditions holds:
-
(a)
f is geodesic radially continuous;
-
(b)
h is odd in the second argument.
If f is geodesic h-pseudolinear, then for any \(x, y\in K\), we have
Theorem 2
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous in the second argument such that the condition (4) holds. Assume that one of the following conditions holds:
-
(a)
f is geodesic radially continuous;
-
(b)
h is odd in the second argument.
Then, the solution set S̄ of OP (3) is geodesic convex.
Proof
Let \(x, y\in \bar{S}\). Then, \(f(x)=f(y)\). By Theorem 1, we have
For any \(s\in [0, 1]\), let \(w:=\exp _{y}s\exp _{y}^{-1}x\). Then,
Again, by Theorem 1 and (5), we obtain
Therefore, \(w\in \bar{S}\) and hence S̄ is geodesic convex. □
For \(\bar{x}\in \bar{S}\), consider the sets
Theorem 3
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous in the second argument such that the condition (4) holds. Assume that one of the following conditions holds:
-
(a)
f is geodesic radially continuous;
-
(b)
h is odd in the second argument.
If \(\bar{x}\in \bar{S}\), then \(\bar{S}=\bar{S}_{1}=\bar{S}_{2}\).
Proof
Since \(x\in \bar{S}\) if and only if \(f(x)=f(\bar{x})\), by Theorem 1, we have \(f(x)=f(\bar{x})\) if and only if \(h(x; \exp _{x}^{-1}\bar{x})=0\). Therefore, \(\bar{S}=\bar{S}_{1}\). Similarly, \(\bar{S}=\bar{S}_{2}\). □
For any \(\bar{x}\in \bar{S}\), consider the set
where \(w=\exp _{\bar{x}}t\exp _{\bar{x}}^{-1}x\) for all \(t\in [0, 1]\).
Theorem 4
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous as well as odd in the second argument such that condition (4) holds. If \(\bar{x}\in \bar{S}\), then \(\bar{S}=\bar{S}_{3}\).
Proof
Let \(x\in \bar{S}\). By Theorem 2, \(w=\exp _{\bar{x}}s\exp _{\bar{x}}^{-1}x\in \bar{S}\) for all \(s\in (0, 1)\). By Theorem 3, we have
As \(x\in \bar{S}\), by Theorem 3 and the oddness of h in the second argument, we have
By combining (6) and (7), we obtain \(\bar{S}\subseteq \bar{S}_{3}\).
Conversely, assume that \(x\in \bar{S}_{3}\), and taking \(t=1\) in particular, we obtain
Therefore, by Theorem 3, we have \(x\in \bar{S}\), and hence, \(\bar{S}=\bar{S}_{3}\). □
For any \(\bar{x}\in \bar{S}\), consider the sets
Theorem 5
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous in the second argument such that condition (4) holds. Assume that one of the following conditions holds:
-
(a)
f is geodesic radially continuous;
-
(b)
h is odd in the second argument.
If \(\bar{x}\in \bar{S}\), then \(\bar{S}=\bar{S}_{4}=\bar{S}_{5}\).
Proof
By Theorem 3, we have \(\bar{S}\subseteq \bar{S}_{4}\). For the converse, assume that \(x\in \bar{S}_{4}\), that is,
Since f is geodesic h-pseudolinear, we have
However, x̄ is a solution of OP (3), and we have \(f(\bar{x})\leq f(x)\). Therefore, \(f(x)=f(\bar{x})\), which yields \(x\in \bar{S}\), and hence, \(\bar{S}=\bar{S}_{4}\).
In a similar way, we obtain \(\bar{S}=\bar{S}_{5}\). □
For any \(\bar{x}\in \bar{S}\), consider the set
where \(w=\exp _{\bar{x}}t\exp _{\bar{x}}^{-1}x\) for all \(t\in [0, 1]\).
Theorem 6
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear and \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous as well as odd in the second argument such that the condition (4) holds. If \(\bar{x}\in \bar{S}\), then \(\bar{S}=\bar{S}_{6}\).
Proof
Since \(\bar{S}_{6}\subseteq \bar{S}_{4}=\bar{S}\). Let \(x\in \bar{S}\), then by Theorem 4, we have
where \(w=\exp _{\bar{x}}t\exp _{\bar{x}}^{-1}x\) for all \(t\in [0, 1]\). Thus, \(x\in \bar{S}_{6}\), and hence, \(\bar{S}\subseteq \bar{S}_{6}\). Therefore, \(\bar{S}=\bar{S}_{6}\). □
For any \(\bar{x}\in \bar{S}\), consider the sets
Theorem 7
Let \(f:K\rightarrow \mathbb{R}\) be geodesic h-pseudolinear. Let \(h:K\times TM\rightarrow \mathbb{R}\cup \{\pm \infty \}\) be positively homogeneous in the second argument such that the condition (4) holds. Assume that one of the following conditions holds:
-
(a)
f is geodesic radially continuous;
-
(b)
h is odd in the second argument.
If \(\bar{x}\in \bar{S}\), then \(\bar{S}=\bar{S}_{7}=\bar{S}_{8}\).
Proof
By Theorem 3, we have \(\bar{S}\subseteq \bar{S}_{7}\). Now, we show that \(\bar{S}_{8}\subseteq \bar{S}\). Let \(x\in \bar{S}_{8}\), then,
Assume on the contrary that \(x\notin \bar{S}\). Then, \(f(x)>f(\bar{x})\). Since f is geodesic h-pseudoconcave, we have
By (8), we have
By the geodesic h-pseudoconvexity of f, we have \(f(x)\leq f(\bar{x})\), which contradicts \(f(x)>f(\bar{x})\). Therefore \(x\in \bar{S}\), and hence, \(\bar{S}_{8}\subseteq \bar{S}\). Thus, \(\bar{S}\subseteq \bar{S}_{7}\subseteq \bar{S}_{8}\subseteq \bar{S}\), and hence, \(\bar{S}=\bar{S}_{7}=\bar{S}_{8}\). □
Remark 1
Theorems 2–7 extend the Theorems 7.1–7.6 in [10], respectively, from Euclidean space settings to Hadamard manifolds.
5 Relations among \(S^{*}\), \(S_{*}\), \(\Gamma (x)\) and \(\Lambda (x)\)
In the present section, we study the relationships among the solution set of NVIP (1) and MTNVIP (2) and sets \(\Gamma (x)\) and \(\Lambda (x)\).
The following proposition follows directly from the definitions.
Proposition 3
Let \(\bar{x}\in K\). Then,
-
(a)
\(\bar{x}\in S^{*} \Leftrightarrow \varphi (\bar{x})=0 \Leftrightarrow \bar{x}\in \Gamma (\bar{x})\);
-
(b)
\(\bar{x}\in S_{*} \Leftrightarrow \Phi (\bar{x})=0 \Leftrightarrow \bar{x}\in \Lambda (\bar{x})\).
Proposition 4
Let \(\bar{x}, \bar{y}\in K\). Then, the following statements are equivalent:
-
(a)
\(\bar{x}\in S^{*}\) and \(\bar{y}\in S_{*}\);
-
(b)
\(h(y; \exp _{y}^{-1}\bar{y})\leq h(\bar{x}; \exp _{\bar{x}}^{-1} \bar{y})\leq h(\bar{x}; \exp _{\bar{x}}^{-1}x), \textit{ for all } x, y \in K\).
Proof
(a) ⇒ (b): Let \(\bar{x}\in S^{*}\) and \(\bar{y}\in S_{*}\). Then, for all \(x, y\in K\), we have
In particular, taking \(x=\bar{y}\), we have
and for \(y=\bar{x}\), we have
Therefore,
Thus, from the definitions of \(S^{*}\) and \(S_{*}\), we obtain the required result.
(b) ⇒ (a): Letting \(y=\bar{y}\) and \(x=\bar{x}\) in the condition (b), we obtain
Therefore, (a) is obtained. □
Proposition 5
The following statements hold:
-
(a)
\(S_{*}\subseteq \Gamma (\bar{x})\), for all \(\bar{x}\in S^{*}\);
-
(b)
\(S^{*}\subseteq \Lambda (\bar{y})\), for all \(\bar{y}\in S_{*}\).
In particular, if h is geodesic upper sign continuous and positively homogeneous in the second argument, then \(S_{*}\subseteq \Gamma (\bar{x})\) for all \(\bar{x}\in S_{*}\). If h is pseudomonotone on \(S^{*}\), then \(S^{*}\subseteq \Lambda (\bar{x})\) for all \(\bar{x}\in S^{*}\).
Proof
By using the proof of Proposition 4, we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\) for all \(\bar{x}\in S^{*}\) and \(\bar{y}\in S_{*}\). Thus, \(\bar{y}\in \Gamma (\bar{x})\) and \(\bar{x}\in \Lambda (\bar{y})\), and hence \(S_{*}\subseteq \Gamma (\bar{x})\) and \(S^{*}\subseteq \Lambda (\bar{y})\).
For the conclusion part, by using the Lemma 1 and the parts (a) and (b), we obtain the required result. □
Proposition 6
-
(a)
For all \(\bar{x}\in S^{*}\) and \(\bar{y}\in \Gamma (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}\bigr) = h\bigl( \bar{y}; \exp _{\bar{y}}^{-1} \bar{x}\bigr) \quad\Leftrightarrow\quad h\bigl( \bar{y}; \exp _{\bar{y}}^{-1} \bar{x}\bigr)=0; \end{aligned}$$ -
(b)
For all \(\bar{x}\in S_{*}\) and \(\bar{y}\in \Lambda (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}\bigr)=h\bigl(\bar{y}; \exp _{\bar{y}}^{-1} \bar{x}\bigr)\quad \Leftrightarrow \quad h\bigl(\bar{x}; \exp _{\bar{x}}^{-1} \bar{y}\bigr)=0. \end{aligned}$$
Proof
It follows from
□
Condition B
For any \(x, y, z\in K\), there exists \(\alpha, \beta >0\) such that
Example 1
Consider \(M=(\mathbb{R}_{++}^{n}, X^{-2})\) is a Hadamard manifold with null sectional curvature, where \(X^{-2}=\operatorname{diag} (\frac{1}{x_{1}^{2}}, \ldots, \frac{1}{x_{n}^{2}} )\) is called the Dikin metric (see [28]). In particular, for \(n=1\), we have \(M=(\mathbb{R}_{++}, X^{-2})\). In particular, we take \(h(x; \cdot )=\langle Ax, \cdot \rangle \) for all \(x\in M\), where \(A:M\rightarrow TM\) is a vector field and let \(Ax=1\) for all \(x\in M\). Then, \(\langle Ax, \exp _{x}^{-1}y \rangle = \exp _{x}^{-1} y\). Now, for any two points \(x, y\in M\), we have
For any \(x, y, z \in M\), we can obtain
that is,
This implies that
Therefore, the condition B holds good with \(\alpha =1\) and \(\beta =\frac{x}{z}\).
Proposition 7
Let \(\bar{x}, \bar{y}\in K\) be such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), that is, they satisfy Condition A. Then, the following assertions hold:
-
(a)
If h is subodd in the second argument, that is, for any \(x\in K\) and \(v\in T_{x}M\), \(h(x; v)\geq -h(x; -v)\), then \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\) if and only if \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\).
-
(b)
If Condition Bholds and either \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\) or \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\), then \(\bar{x}\in S^{*}\) if and only if \(\bar{y}\in S^{*}\).
Proof
(a) Suppose that \(h(\bar{x}; \exp _{\bar{x}}^{-1} \bar{y})=0\). Since h is subodd in the second argument, we have
Thus,
Then, both the vectors \(\exp _{\bar{x}}^{-1} \bar{y}\) and \(P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1} \bar{x}\) belong to \(\mathbb{U}(\bar{x})\). Since \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), we have \(\exp _{\bar{x}}^{-1}\bar{y}\in \mathbb{V}(\bar{y})\) and \(P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1} \bar{x} \in \mathbb{V}( \bar{y})\). Therefore,
that is,
and
that is,
Since h is subodd in the second argument, we obtain
Therefore,
From equations (9) and (10), we have
(b) If either \(h(\bar{x}; \exp _{\bar{x}}^{-1} \bar{y})=0\) or \(h(\bar{y}; \exp _{\bar{y}}^{-1} \bar{x})=0\), then it follows from part (a).
Let us suppose that \(h(\bar{x}; \exp _{\bar{x}}^{-1} \bar{y})=0\) and \(\bar{x}\in S^{*}\). Then, \(h(\bar{x}; \exp _{\bar{x}}^{-1} z)\geq 0\) for all \(z\in K\). Therefore, by Condition B, we have \(\alpha, \beta >0\) such that
Since \(\beta >0\), we have \(h(\bar{x}; P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1} z)\geq 0 \text{ for all } z\in K\). Therefore, \(P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1} z\in \mathbb{U}(\bar{x})\). Since \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), we have
that is,
Since h is subodd in the second argument, we have for all \(z\in K\),
that is,
Therefore, \(\bar{y}\in S^{*}\). In a similar manner, the converse part can be proved. □
Theorem 8
Let \(\bar{x}, \bar{y}\in K\) be such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) and assume that the Condition Bholds. Then, the following assertions hold:
-
(a)
\(\bar{x}\in S^{*}\) and \(\bar{y}\in \Gamma (\bar{x})\) if and only if \(\bar{x}\in \Gamma (\bar{x})\) and \(\bar{y}\in S^{*}\);
-
(b)
\(\bar{x}\in S^{*}\) and \(\bar{y}\in \Lambda (\bar{x})\) if and only if \(\bar{x}\in \Lambda (\bar{x})\) and \(\bar{y}\in S^{*}\).
Proof
(a) Let \(\bar{x}\in S^{*}\) and \(\bar{y}\in \Gamma (\bar{x})\). Then, \(\bar{x}\in \Gamma (\bar{x})\) and \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=- \varphi (\bar{x})=0\), and by Proposition 7(b), we obtain \(\bar{y}\in S^{*}\). For the converse part, take \(\bar{x}\in \Gamma (\bar{x})\) and \(\bar{y}\in S^{*}\). Then, for any \(\bar{x}\in S^{*}\), we have
Therefore, \(\exp _{\bar{x}}^{-1}\bar{y}\in \mathbb{U}(\bar{x})\). Since \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), we have
and hence,
However, \(\bar{y}\in S^{*}\), we have
Thus,
and by Proposition 7(a), we have
Therefore, \(\bar{y}\in \Gamma (\bar{x})\).
(b) Since \(\bar{x}\in S^{*}\) and \(\bar{y}\in \Lambda (\bar{x})\), we have
Hence, \(\exp _{\bar{x}}^{-1}\bar{y}\in \mathbb{U}(\bar{x})=\mathbb{V}( \bar{y})\), and thus, \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\leq 0\). Therefore,
and hence \(\Phi (\bar{x})=0\), that is, \(\bar{x}\in \Lambda (\bar{x})\). Since \(\bar{x}\in S^{*}\) and by Proposition 7(b), we have \(\bar{y}\in S^{*}\).
For the converse part, since \(\bar{x}\in \Lambda (\bar{x})\), we have \(\Phi (\bar{x})=0\), and hence \(\bar{x}\in S_{*}\). Thus, by Proposition 5(b), we obtain \(\bar{y}\in \Lambda (\bar{x})\) and by part (a) of Proposition 5, we have \(\bar{x}\in \Gamma (\bar{y})\). Therefore, by part (a), we obtain \(\bar{x}\in S^{*}\). □
Remark 2
The following statements also hold under the assumptions of Theorem 8:
-
(a)
\(\bar{y}\in S^{*}\) and \(\bar{x}\in \Gamma (\bar{y})\) if and only if \(\bar{y}\in \Gamma (\bar{y})\) and \(\bar{x}\in S^{*}\);
-
(b)
\(\bar{y}\in S^{*}\) and \(\bar{x}\in \Lambda (\bar{y})\) if and only if \(\bar{y}\in \Lambda (\bar{y})\) and \(\bar{x}\in S^{*}\).
Remark 3
If h is subodd in the second argument. Then, for any \(\bar{x}, \bar{y}\in S^{*}\) such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), we have
Proposition 8
Let \(k<0\) and h be subodd in the second argument, and \(\bar{x}, \bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x}) \cup S^{*} \cup \Gamma (\bar{y})\cup \Lambda (\bar{y})\). Consider the following statements
-
(a)
\(h(\bar{y}; P_{\bar{y}, \bar{x}}v)=k h(\bar{x}; v)\), for all \(v\in T_{\bar{x}}M\);
-
(b)
\(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\);
-
(c)
\(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})\leq 0 \Leftrightarrow h( \bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\geq 0\), and \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})\geq 0 \Leftrightarrow h( \bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\leq 0\);
-
(d)
\(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})= 0\) and \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})= 0\);
-
(e)
\(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}) = h(\bar{y}; \exp _{\bar{y}}^{-1} \bar{x})\).
Then, (a) ⇒ (b) ⇒ (c) ⇒ (d) ⇒ (e). Furthermore, if (d), (e) ⇒ (a), then (a) ⇔ (b) ⇔ (c) ⇔ (d) (⇔ (e)).
Proof
The following implications are trivial. (a) ⇒ (b) ⇒ (c) and (d) ⇒ (e). Now, to show (c) ⇒ (d), let us suppose that (c) holds for some \(\bar{x}, \bar{y}\in K\). Therefore, if
then (d) follows. To complete the proof, we assert that if (c) is satisfied for \(\bar{x}, \bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x})\cup S^{*} \cup \Gamma (\bar{y})\cup \Lambda (\bar{y})\), then we can consider (11) without loss of generality. Indeed, if \(\bar{x}\in \Gamma (\bar{x})\cup S^{*}\cup \Lambda (\bar{y})\), then \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})\geq 0\). On the other hand, if \(\bar{x}\in \Lambda (\bar{x})\cup \Gamma (\bar{y})\), then \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\leq 0\), and hence by (c), we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})\geq 0\). Now, If \(\bar{y}\in S^{*}\cup \Gamma (\bar{y})\cup \Lambda (\bar{x})\), we obtain \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\geq 0\), and if \(\bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{y})\), we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})\leq 0\), by (c), which implies that \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})\geq 0\). □
Definition 9
A bifunction \(h:K\times TM \rightarrow \mathbb{R}\cup \{\pm \infty \}\) is said to be pseudomonotone∗ if it is pseudomonotone and for some \(k <0\) and for all \(x, y\in K\),
Remark 4
Since for \(\bar{x}\in S_{*}\) and \(\bar{y}\in \Lambda (\bar{x})\), we have \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\), assertions (d) and (e) in Proposition 8 are equivalent to \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\). Therefore, if we add the assumption pseudomonotone∗ of h in the Proposition 8, then (a) to (e) in Proposition 8 are equivalent to each other.
6 Minimum principle sufficiency property for nonsmooth variational inequalities
The NVIP (1) has the minimum principle sufficiency property if
Theorem 9
Suppose that the Condition Bholds. Then, the following statements hold:
-
(a)
If \(\bar{x}\in S^{*}\) and \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{y}\in \Gamma (\bar{x})\), then \(\Gamma (\bar{x})\subseteq S^{*}\);
-
(b)
If \(\bar{x}\in S_{*}\) with \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{y}\in \Gamma (\bar{x})\), then \(\bar{x}\in \Gamma (\bar{x})=S^{*}\);
-
(c)
If \(\bar{x}\in S^{*}\cup S_{*}\) and \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{y}\in S^{*}\), then \(\bar{x}\in S^{*}\subseteq \Gamma (\bar{x})\).
Proof
(a) Follows directly by using the part (a) of Theorem 8.
(b) For \(\bar{x}\in S_{*}\), let \(\bar{y}\in \Gamma (\bar{x})\) with \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\). Then,
Since h is subodd in the second argument, we have
and hence
Therefore, \(P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1}\bar{x}\in \mathbb{U}( \bar{x})\). Since \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), we have
and hence
By the suboddness of h in the second argument, we obtain
Therefore,
Since \(\bar{x}\in S_{*}\), we have
and hence
Therefore by Proposition 6(a), we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})= 0\). Now, on using (12), we have \(\varphi (\bar{x})=0\), which means that \(\bar{x}\in S^{*}\). Hence, by Proposition 7(b), \(\bar{y}\in S^{*}\).
On the other hand, take \(\bar{y}\in S^{*}\), then
By Proposition 7(b), we have \(\bar{x}\in S^{*}\), which implies that \(\bar{x}\in \Gamma (\bar{x})\). By part (a) of the Proposition 7, we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\). Therefore, \(\bar{y}\in \Gamma (\bar{x})\) and hence \(S^{*}\subseteq \Gamma (\bar{x})\).
(c) Let \(\bar{y}\in S^{*}\). If \(\bar{x}\in S^{*}\), then from part (a) of the Theorem 8, we obtain \(\bar{x}\in \Gamma (\bar{x})\), and hence, \(S^{*}\subseteq \Gamma (\bar{x})\). If \(\bar{x}\in S_{*}\), then,
that is, \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\). Now, from Proposition 7(b), we obtain \(\bar{x}\in S^{*}\) and by Proposition 7(a), we obtain \(h(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y})=0\), which implies that \(\bar{y}\in \Gamma (\bar{x})\), and hence \(S^{*}\subseteq \Gamma (\bar{x})\). □
Remark 5
By Proposition 5 and Theorem 9, the NVIP (1) has the minimum sufficiency property if \(S^{*}\subseteq S_{*}\) and Condition A holds for all \(\bar{x}\in S^{*}\) and for all \(\bar{y}\in \Gamma (\bar{x})\) and Condition B holds.
7 Maximum principle sufficiency property for nonsmooth variational inequalities
The NVIP (1) has the maximum principle sufficiency property if
Theorem 10
Suppose that the Conditions Aand Bhold. Then, the following statements hold:
-
(a)
If \(\bar{x}\in S^{*}\cup S_{*}\) and \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{y}\in \Lambda (\bar{x})\), then \(\bar{x}\in \Lambda (\bar{x})=S^{*}\);
-
(b)
If for each \(\bar{x}\in S^{*}\), there exists \(\bar{y}\in \Lambda (\bar{x})\) such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), then \(S^{*}\subseteq S_{*}\);
-
(c)
If for each \(\bar{x}\in S_{*}\), there exists \(\bar{y}\in S^{*}\) such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), then \(S_{*}\subseteq S^{*}\);
-
(d)
If \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\), \(S^{*}=\Lambda (\bar{x})=S_{*}\) for all \(\bar{x}\in S^{*} \cup S_{*}\).
Proof
(a) Suppose that \(\bar{x}\in S^{*}\cup S_{*}\) and \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{y}\in \Lambda (\bar{x})\). If \(\bar{x}\in S^{*}\), then by part (b) of Theorem 8, we obtain \(\bar{x}\in \Lambda (\bar{x})\) and hence \(\Lambda (\bar{x})\subseteq S^{*}\). On the other hand, if \(\bar{x}\in S_{*}\), then Proposition 3(b), \(\bar{x}\in \Lambda (\bar{x})\) and by Proposition 5(b), we obtain \(S^{*}\subseteq \Lambda (\bar{x})\), which implies that
Therefore, the above equality holds in particular for all \(\bar{z}\in S^{*}\); thus, by Proposition 7(b), \(\bar{x}\in S^{*}\), and again by Proposition 7(b), \(\bar{y}\in S^{*}\). Therefore, \(\Lambda (\bar{x})=S^{*}\).
(b) It directly follows from Theorem 8.
(c) By Proposition 5(b), we have \(S^{*}\subseteq \Lambda (\bar{x})\), for all \(\bar{x}\in S_{*}\). Therefore, if \(\bar{x}\in S_{*}\), then there exists \(\bar{y}\in S^{*}\subseteq \Lambda (\bar{x})\) such that \(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\), which implies that \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\) and by Proposition 7(b), we obtain \(\bar{x}\in S^{*}\) since \(\bar{y}\in S^{*}\).
(d) Assume that Condition A holds for all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\). Then from (a), we obtain
Therefore, from (a) and (b), we have \(S^{*}=S_{*}\). □
Theorem 11
Assume that
for some \(k<0\) and for all \(v\in T_{\bar{x}}M\), \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\). Then, the following statements are equivalent:
-
(a)
\(\mathbb{U}(\bar{x})=\mathbb{V}(\bar{y})\) for all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\);
-
(b)
\(S^{*}=\Lambda (\bar{x})=S_{*}\) for all \(\bar{x}\in S^{*}\cup S_{*}\);
-
(c)
For all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}\bigr) \geq 0 \quad \Leftrightarrow \quad h\bigl(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x}\bigr) \leq 0, \end{aligned}$$and
$$\begin{aligned} h\bigl(\bar{x}; P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1}\bar{x}\bigr) \geq 0 \quad\Leftrightarrow\quad h\bigl(\bar{y}; P_{\bar{y}, \bar{x}}\exp _{ \bar{x}}^{-1} \bar{y}\bigr) \leq 0; \end{aligned}$$ -
(d)
For all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}\bigr)=0 \quad\textit{and} \quad h \bigl( \bar{y}; \exp _{\bar{y}}^{-1}\bar{x}\bigr)=0; \end{aligned}$$ -
(e)
\(h(\bar{y}; P_{\bar{y}, \bar{x}}v) = k h(\bar{x}; v)\) for all \(v\in T_{\bar{x}}M\), for some \(k<0\), all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Lambda (\bar{x})\).
Proof
By Proposition 8, we obtain the equivalence among (a), (c), (d) and (e). Now, from Theorem 10(d), we have (a) ⇒ (b) and lastly from Proposition 4, we obtain (b) ⇒ (d). □
Remark 6
In the case \(S^{*}\subseteq S_{*}\), the expression \(S^{*}\cup S_{*}\) in Theorem 11 can be reduced to \(S_{*}\) and the equality \(h(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x})=0\) in (d) can also be omitted.
Under geodesic upper sign continuity and pseudomonotonicity∗ of h, all the statements of Theorem 11 hold, as shown below.
Theorem 12
Let h be a geodesic upper sign continuous and pseudomonotone∗. Then, for some \(k>0\), (a)–(e) in the Theorem 11hold.
Proof
If h is geodesic upper sign continuous and pseudomonotone, then \(S^{*}= S_{*}\). Since (d) can be easily verified for \(\bar{x}\in S_{*}\) and \(\bar{y}\in \Lambda (\bar{x})\), then by Theorem 11, (a)–(e) hold. □
Theorem 13
For some \(k>0\), the following statements satisfy:
-
(a)
The Condition Aholds for all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Gamma (\bar{x}) \cup \Lambda (\bar{x})\);
-
(b)
\(S^{*}=\Gamma (\bar{x})=\Lambda (\bar{x})=S_{*}\) for all \(\bar{x}\in S^{*}\cup S_{*}\);
-
(c)
For all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1}\bar{x}\bigr) \geq 0 \quad\Leftrightarrow\quad h\bigl(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x} \bigr) \geq 0, \end{aligned}$$and
$$\begin{aligned} h\bigl(\bar{x}; \exp _{\bar{x}}^{-1}\bar{y}\bigr)\geq 0\quad \Leftrightarrow\quad h\bigl(\bar{y}; P_{\bar{y}, \bar{x}}\exp _{\bar{x}}^{-1} \bar{y}\bigr) \geq 0; \end{aligned}$$ -
(d)
For all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x})\),
$$\begin{aligned} h\bigl(\bar{x}; P_{\bar{x}, \bar{y}}\exp _{\bar{y}}^{-1}\bar{x} \bigr)=0 \quad\textit{and}\quad h\bigl(\bar{y}; \exp _{\bar{y}}^{-1}\bar{x} \bigr)=0; \end{aligned}$$ -
(e)
\(h(\bar{y}; P_{\bar{y}, \bar{x}}v)= k h(\bar{x}; v)\) for all \(v\in T_{\bar{x}}M\), for some \(k>0\), all \(\bar{x}\in S^{*}\cup S_{*}\) and all \(\bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x})\).
Hence, if (e) ⇒ (d), then (a)–(e) are equivalent.
Proof
By Proposition 8, we have (e) ⇒ (a) ⇒ (c) ⇒ (d) and from Proposition 3(b) and Theorem 10(d), we obtain (a) ⇒ (b). Now, we assert that (b) ⇒ (d). Indeed, let (b) hold. Then, for each \(\bar{x}\in S^{*}\cup S_{*}\), we have \(\bar{x}\in S^{*}=S_{*}\) and \(\Gamma (\bar{x})\cup \Lambda (\bar{x})=\Gamma (\bar{x}) \cap \Lambda (\bar{x})\). Thus, for each \(\bar{y}\in \Gamma (\bar{x})\cup \Lambda (\bar{x})\), (d) holds. □
8 Error bounds for nonsmooth variational inequalities
Definition 10
A bifunction \(h:K\times TM \rightarrow \mathbb{R}\cup \{\pm \infty \}\) is said to be strongly pseudomonotone if for all \(x, y\in K\), there exists \(\alpha >0\) such that
where \(d(\cdot, \cdot)\) is the Riemannian distance.
Theorem 14
Assume that h is odd in the second argument and strongly pseudomonotone with constant \(\alpha >0\). Then,
where \(d(x, S^{*})=\inf_{y\in S^{*}} d(x, y)\) is the Riemannian distance between the point x and the solution set \(S^{*}\).
Proof
Let \(\bar{x}\in S^{*}\), then
Since h is strongly pseudomonotone, there exists \(\alpha >0\) such that
As h is odd in the second argument and \(-\exp _{y}^{-1}x=P_{y, x}\exp _{x}^{-1}y\), we have
and hence,
Since
we have
Therefore,
the desired result. □
The NVIP (1) has weak sharp solutions if
Therefore, if h is odd in the second argument and strongly pseudomonotone with constant \(\alpha >0\), then by Theorem 14, the NVIP (1) has weak sharp solutions.
9 Conclusion
In this paper, we define the notions minimum and maximum principle sufficiency properties for nonsmooth variational inequalities by using gap functions in the setting of Hadamard manifolds. We also characterize these two sufficiency properties in the setting of Hadamard manifolds. We conclude our paper by introducing the idea of the error bounds for nonsmooth variational inequalities in the setting of Hadamard manifolds. The main objective of the paper is to include the existing results in nonlinear space, namely Hadamard manifolds. This is the first paper dealing with these two notions for nonsmooth variational inequalities. Therefore, in the future one can extend the concept of this paper in many other directions.
Availability of data and materials
Not applicable.
References
Agarwal, P., Dragomir, S.S., Jleli, M., Samet, B. (eds.): Advances in Mathematical Inequalities and Applications Springer, Berlin (2018)
Agarwal, P., Jleli, M., Tomar, M.: Certain Hermite–Hadamard type inequalities via generalized k-fractional integrals. J. Inequal. Appl. 2017, 55 (2017). https://doi.org/10.1186/s13660-017-1318-y
Al-Homidan, S., Ansari, Q.H., Islam, M.: Browder type fixed point theorem on Hadamard manifolds with applications. J. Nonlinear Convex Anal. 20(11), 2397–2409 (2019)
Al-Homidan, S., Ansari, Q.H., Islam, M.: Existence results and proximal point algorithm for equilibrium problems on Hadamard manifolds. Carpath. J. Math. 37(3), 393–406 (2021)
Al-Homidan, S., Ansari, Q.H., Nguyen, L.V.: Weak sharp solutions for nonsmooth variational inequalities. J. Optim. Theory Appl. 175(3), 683–701 (2017)
Alshahrani, M., Al-Homidan, S., Ansari, Q.H.: Minimum and maximum principle sufficiency properties for nonsmooth variational inequalities. Optim. Lett. 10(4), 805–819 (2016)
Ansari, Q.H., Islam, M.: Explicit iterative algorithms for solving equilibrium problems on Hadamard manifolds. J. Nonlinear Convex Anal. 21(2), 425–439 (2020)
Ansari, Q.H., Islam, M., Yao, J.-C.: Nonsmooth variational inequalities on Hadamard manifolds. Appl. Anal. 99(2), 340–358 (2020)
Ansari, Q.H., Islam, M., Yao, J.C.: Nonsmooth convexity and monotonicity in terms of a bifunction on Riemannian manifolds. J. Nonlinear Convex Anal. 18(4), 743–762 (2017)
Ansari, Q.H., Lalitha, C.S., Mehta, M.: Generalized Convexity, Nonsmooth Variational Inequalities, and Nonsmooth Optimization. CRC Press, Boca Raton (2014)
Artsawang, N., Ungchittrakool, K.: Characterizations of minimum and maximum properties for generalized nonsmooth variational inequalities. J. Nonlinear Convex Anal. 19, 731–748 (2018)
Bertsekas, D.P.: Necessary and sufficient conditions for a penalty function to be exact. Math. Program. 9, 8–99 (1975)
Burke, J.V., Ferris, M.C.: Weak sharp minima in mathematical programming. SIAM J. Control Optim. 31, 1340–1359 (1993)
Chavel, I.: Riemannian Geometry- A Mordern Introduction. Cambridge University Press, London (1993)
Colao, V., López, G., Marino, G., Martín-Márquez, V.: Equilibrium problems in Hadamard manifolds. J. Math. Anal. Appl. 388, 61–77 (2012)
Docarmo, M.P.: Riemannian Geometry. Birkhäuser, Boston (1992)
Ferris, M.C.: Weak sharp minima and penalty functions in mathematical programming. Ph.D. Thesis, University of Cambridge (1988)
Ferris, M.C.: Finite termination of the proximal point algorithm. Math. Program. 50, 359–366 (1991)
Ferris, M.C., Managasarian, O.L.: Minimum principle sufficiency. Math. Program. 57, 1–14 (1992)
Hartman, P., Stampacchia, G.: On some non-linear elliptic differential-functional equations. Acta Math. 115, 271–310 (1966)
Inoan, D.I.: Existence results for systems of quasi-variational relations. Constr. Math. Anal. 2(4), 217–222 (2019)
Kadakal, M., İşcan, İ., Agarwal, P., Jleli, M.: Exponential trigonometric convex functions and Hermite–Hadamard type inequalities. Math. Slovaca 71(1), 43–56 (2021). https://doi.org/10.1515/ms-2017-0410
Li, S.L., Li, C., Liou, Y.C., Yao, J.C.: Existence of solutions for variational inequalities on Riemannian manifolds. Nonlinear Anal. 71, 5695–5706 (2009)
Marcotte, P., Zhu, D.: Weak sharp solutions of variational inequalities. SIAM J. Optim. 9, 179–189 (1999)
Németh, S.Z.: Variational inequalities on Hadamard manifolds. Nonlinear Anal. 52, 1491–1498 (2003)
Petersen, P.: Riemannian Geometry. GTM 171, 2nd edn. Springer, Berlin (2006)
Polyak, B.T., Tretiyakov, N.V.: Concerning an iterative methods for linear programming and its economic interpretation. Econ. Math. Methods 8(5), 740–751 (1972) English Transalation: Matekon, 10(3), 81–100 (1974)
Rapcsak, T.: Smooth Nonlinear Optimization in \(\mathbb{R}^{n}\). Kluwer Academic, Dordrecht (1997)
Sakai, T.: Riemannian Geometry. Translations of Mathematical Monographs, vol. 149. Am. Math. Soc., Providence (1992)
Spivak, M.: Calculus on Manifolds: A Modern Aproach to Classical Theorems of Advanced Calculus. Benjamin, New York (1965)
Udriste, C.: Convex Functions and Optimization Methods on Riemannian Manifolds. Mathematics and Its Applications, vol. 297. Kluwer Academic, Dordrecht (1994)
Vivas-Cortez, M., Ali, M.A., Budak, H., Kalsoom, H., Agarwal, P.: Some new Hermite–Hadamard and related inequalities for convex functions via \((p, q)\)-integral. Entropy 23(7), 828 (2021). https://doi.org/10.3390/e23070828
Wu, Z., Lu, Y.: Minimum and maximum principle sufficiency for a nonsmooth variational inequality. Bull. Malays. Math. Sci. Soc. 44, 1233–1257 (2021)
Wu, Z.L.: Minimum principle sufficiency for a variational inequality with pseudomonotone mapping. WSEAS Trans. Math. 16, 48–56 (2017)
Wu, Z.L., Wu, S.Y.: Weak sharp solutions of variational inequalities in Hilbert spaces. SIAM J. Optim. 14(4), 1011–1027 (2004)
You, X.X., Ali, M.A., Budak, H., Agarwal, P., Chu, Y.M.: Extensions of Hermite–Hadamard inequalities for harmonically convex functions via generalized fractional integrals. J. Inequal. Appl. 2021, Article ID 102 (2021). https://doi.org/10.1186/s13660-021-02638-3
Acknowledgements
The authors would like to thank the editors and the anonymous referees for the detailed and valuable suggestions that helped to improve the original manuscript to its present form.
Funding
There is no funding for this paper.
Author information
Authors and Affiliations
Contributions
The authors read and approved the final manuscript. All authors contributed equally to writing this paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare that they have no competing interests.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Islam, M., Irfan, S.S. Some aspects of nonsmooth variational inequalities on Hadamard manifolds. J Inequal Appl 2021, 198 (2021). https://doi.org/10.1186/s13660-021-02733-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s13660-021-02733-5