Jump to ContentJump to Main Navigation
Show Summary Details

Open Mathematics

formerly Central European Journal of Mathematics

Editor-in-Chief: Gianazza, Ugo / Vespri, Vincenzo

1 Issue per year


IMPACT FACTOR 2015: 0.512

SCImago Journal Rank (SJR) 2015: 0.521
Source Normalized Impact per Paper (SNIP) 2015: 1.233
Impact per Publication (IPP) 2015: 0.546

Mathematical Citation Quotient (MCQ) 2015: 0.39

Open Access
Online
ISSN
2391-5455
See all formats and pricing
Volume 14, Issue 1 (Jan 2016)

Issues

The fuzzy metric space based on fuzzy measure

Jialiang Xie
  • College of Science, Jimei University, Xiamen, Fujian, 361021, P. R. China
  • Email:
/ Qingguo Li
  • Corresponding author
  • College of Mathematics and Econometrics, Hunan University, Changsha, Hunan, 410082, P. R. China
  • Email:
/ Shuili Chen
  • Chengyi University College, Jimei University, Xiamen, Fujian, 361021, P. R. China
/ Huan Huang
  • College of Science, Jimei University, Xiamen, Fujian, 361021, P. R. China
Published Online: 2016-09-20 | DOI: https://doi.org/10.1515/math-2016-0051

Abstract

In this paper, we study the relation between a fuzzy measure and a fuzzy metric which is induced by the fuzzy measure. We also discuss some basic properties of the constructed fuzzy metric space. In particular, we show that the nonatom of fuzzy measure space can be characterized in the constructed fuzzy metric space.

Keywords: Fuzzy measure; Fuzzy metric space; t-Norm, Nonatom; Convex

MSC 2010: 54E70; 54A40; 60B05

1 Introduction

The theory of metric space is an important topic in topology. The methods of constructing a fuzzy metric have been extensively studied [14]. It is worth noting that George and Veeramani [5] introduced the concept of a fuzzy metric with the help of continuous t-norms. Despite being restrictive, this kind of fuzzy metric provides a more natural and intuitive way to connect with the metrizable topological spaces. This concept has been widely used in various papers devoted to fuzzy topology [511]. It also has been applied to color image filtering to improve image quality (see [9] and the references therein).

On the other hand, measure theory is one of the most important theories in mathematics and it has been extensively studied. The concept of fuzzy measure was first introduced by Sugeno [12]. It can be regarded as an extension of classical measure in which the additivity is replaced by a weaker condition, monotonicity. Klement et. al establish the axiomatic theory of fuzzy σ-algebras and develop a measure theory of fuzzy sets [1315]. So far, there are many different classes of fuzzy measures, such as possibility measure [16, 17], decomposable measure [1820], pseudo-additive measure [21, 22], and T-measure [2326] etc. A systematic study of fuzzy measure theory can be found in [2730].

Recently, the study of constructing a fuzzy metric using a fuzzy measure technique has been actively pursued. In particular, a fuzzy Prokhorov metric and ultrametric defined on the set of all probability measures in a compact fuzzy metric space have been developed in [31, 32]. Cao et. al [33] introduce fuzzy analogue of the Kantorovich metric among the set of possibility distributions. In [34, 35], the authors discuss the relations between the decomposable measure and the fuzzy metric. More specifically, it has been proven that, with a Hausdorff fuzzy pseudo-metric constructed on its power set, a stationary fuzzy ultrametric space can induce a σ-⊥-superdecomposable measure. The authors of [35] further use a topological approach to extend the t-conorm-based decomposable measures by introducing a fuzzy pseudometric structure on an algebra of sets. In [36] we constructed a pseudo-metric (in the sense of Pap) on the measurable sets of a given σ-⊥-decomposable measure, and then analyzed the connection between the induced pseudo-metric and the σ-⊥-decomposable measure.

In this paper we focus on the following problems: how to construct a fuzzy metric by using a fuzzy measure developed in [14, 15] and what is the relation between these two? And what is the relations between these two? Specifically, by introducing the concept of an equivalence relation on fuzzy measurable sets, we construct a fuzzy metric on the associated quotient sets from a given fuzzy measure. Furthermore, we study some basic properties of the constructed fuzzy metric space such as completeness and continuity. To gain better insight into our proposed method of constructing a fuzzy metric, we study the properties of the constructed fuzzy metric which can precisely reflect those of fuzzy measure. As an illustration we obtain that the nonatom of fuzzy measure space can be characterized in the constructed fuzzy metric space.

The rest of the paper is organized as follows. In Section 2, some basic notions and results are given. Sections 3 and 4 are devoted to constructing a fuzzy metric and discussing its properties. In Section 5, we discuss the relationships between the constructed fuzzy metric and the fuzzy measure. Finally, some concluding remarks are given in Section 6.

2 Preliminaries

We start this section by recalling the concept of triangular norms from [20, 37]. They are an important tool in extending the classical metric space to fuzzy metric space.

Definition 2.1 (Klement et al.): A function ⊤ : [0, 1]2 → [0, 1] is called triangular norm (t-norm for short) if it satisfies the following conditions for all a, b, c, d ∈ [0, 1]:

  • (T1)

    a⊤1 = a. (boundary condition)

  • (T2)

    abcd whenever ac and bd.(monotonicity)

  • (T3)

    ab = bb. (commutativity)

  • (T4)

    a⊤(bc) = (ab)⊤c. (associativity)

A t-Norm ⊤ is said to be continuous if it is a continuous function in [0, 1]2. Typical examples of continuous t-Norms are the minimum TM, the product TP and the Łukasiewicz t-norm TL, which are given by, respectively: TM(x, y)=min(x, y), TP(x, y)=xy, TL(x, y)=max(0, x+y1). Because of the associative property, the t-Norm ⊤ can be extended by induction to n-ary operation by setting i=1nxi=(i=1n1xi)xn Due to monotonicity, for each sequence (xi)i∈ℕ of elements of [0, 1], the following limit can be considered: i=1xi=limni=1nxi. Next we recall the concept of a fuzzy metric with the help of the continuous t-norm, which is a generalization of the concept of Menger probabilistic metric to the fuzzy setting.

Definition 2.2 (George and Veeramani [5]): The 3-tuple (X, M, ⊤) is said to be a fuzzy metric space if X is an arbitrary nonempty set, ⊤ is a continuous t-norm and M is a fuzzy set on X2 × (0, ∞) satisfying the following conditions, for all x, y, z ∈ X, t, s > 0:

  • (GV1)

    M(x, y, t) > 0,

  • (GV2)

    M(x, y, t) = 1 iff x = y,

  • (GV3)

    M(x, y, t) = M(x, y, t)

  • (GV4)

    M(x, z, t + s) ≥ M(x, y, t)⊤M(y, z, t)

  • (GV5)

    M(x, y, ·) : (0, ∞) → (0, 1] is continuous.

If the condition (GV2) is replaced by the condition (GVp): M(x, x, t) = 1, then (X, M, ⊤) is said to be a fuzzy pseudometric space.It was proved in [5] that in a fuzzy metric space X, the function M(x, y, ·) is nondecreasing for all x, yX. A sequence (xi)i∈ℕ in a fuzzy metric space (X, M, ⊤) is said to converge [6] to x if limi→∞ M(xi, x, t) = 1 for all t > 0; a sequence (xi)i∈ℕ in a fuzzy metric space (X, M, ⊤) is said to be Cauchy [6] if limi, j →∞ M(xi, x, t) = 1 for all t > 0; (X, M, ⊤) is said to be complete [8] if every Cauchy sequence is convergent. A mapping f from a fuzzy metric space (X, M, ⊤1) to a fuzzy metric space (X, M, ⊤2) is called uniformly continuous [7] if for each ε ∈ (0, 1) and each t > 0, their exist r ∈ (0, 1) and s > 0 such that N(f(x), f(y), t) > 1 – ε whenever M(x, y, s) > 1 – r.

In this follow-up, we recall several concepts from the measure theory of fuzzy sets (see e.g. [1315]).

Definition 2.3: Let X be a nonempty set, I the unit interval [0, 1]. A subset A of IX is a fuzzy σ-algebra iff

  • (A1)

    0, 1A,

  • (A2)

    AA implies 1 — A ∈ A,

  • (A3)

    if {Ai}i=1 is a sequence in A, then i=1Ai = sup AiA.

Definition 2.4: A finite fuzzy measure (or F-measure) on a fuzzy σ-algebra A is a function μ : A → [0, ∞) satisfying:

  • (M1)

    μ(0) = 0,

  • (M2)

    for AA, μ(1 – A) = μ(1) – μ(A),

  • (M3)

    for A, BA, μ(A ∨ B) + μ (AB) = μ(A) + μ(B),

  • (M4)

    if {Ai}i=1 is a sequence in A such that AiA, AA, then μ(A) sup = μ(Ai).

We call (X, A, μ) an F-measure space, elements of A are referred as fuzzy measurable sets.

3 Constructing fuzzy metric based on F-measure

In this follow-up, ⊤ stands for the minimum t-norm TM. The following result is the natural fuzzy metric structure on fuzzy measurable sets.

Theorem 3.1: Let (X, A, μ) be an F-measure space. If we define the fuzzy set M on A2 × (0, ∞) by M(A, B, t)=tt+μ(AB)μ(AB), where A, B ∈ A. Then M is a fuzzy pseudometric on A.

Proof: From the definition of M, it is obvious that for any A, BA, t > 0, we have (i) M(A, B, t) > 0; (ii) M(A, A, t) = 1 and M(A, B, t) = M(B, A, t); (iii) M(x, y, ·) : (0, ∞) → [0, 1] is continuous. The only thing that we need to prove is the triangular inequality. For any A, B, C ∈ A and t, s > 0, we have M(A, B, t)=tt+μ(AB)μ(AB), M(B, C, s)=ss+μ(BC)μ(BC) and M(A, C, t+s)=t+st+s+μ(AC)μ(AC). We are going to verify M(A, B, C, t + s) ≥ M(A, B, t) ⊤ M(B, C, s). For each A, B, C, ∈ A, we have μ(AB)+μ(BC)+μ(AC)=μ(ABC)+μ(AB)(BC)+μ(AC)=μ(ABC)+μB(AC)+μ(AC)μ(AC)+μB(AB)C+μ(ABC)=μ(AC)+μ(AB)+μ(BC) and so μ(AC)μ(AC)μ(AB)μ(AB)+μ(BC)μ(BC) we can distinguish two cases: M(B, C, s) ≥ M(A, B, s) or M(A, B, t) ≥ M(B, C, s).Case (i). If M(B, C, s) ≥ M(A, B, t), or equivlently, ss+μ(BC)μ(BC)tt+μ(AB)μ(AB) and hence s(μ(AB)μ(AB))t(μ(BC)μ(BC)) Consequently, t(μ(AC)μ(AC))t(μ(AB)μ(AB))+t(μ(BC)μ(BC))t(μ(AB)μ(AB))+s(μ(AB)μ(AB)). This implies that tt+μ(AB)μ(AB)t+st+s+μ(AC)μ(AC), and hence M(A, C, t+s)M(A, B, t)=M(A, B, t)M(B, C, s). Case (ii). Similar to (i). Thus, M is a fuzzy pesudometric on A.

Remark 3.2: Based on Theorem 3.1, M is a fuzzy pseudometric on A under any left-continuous t-norms since TM is the biggest left-continuous t-norm.

Generally speaking, the fuzzy pseudometric space ( A, M, ⊤) is usually not a fuzzy metric space. But we can construct a fuzzy metric space from a fuzzy pseudometric metric space ( A, M, ⊤) and, at the same time, keep the general characteristics of the fuzzy pseudometric metric space.

Lemma 3.3: Let (X, A, μ) be an F-measure space. For each A, B ∈, we define the relation “~” on A : A ~ B if and only if μ(AB) = μ(AB). Then ~ is an equivalence relation on A.

Theorem 3.4: Let (X, A, μ) be an F-measure space. Let A/μ be the set of all equivalence classes for the relation “∼”. The fuzzy pseudometric M has a natural extension to M([A], [B], t)=M(A, B, t), where [A] ([B]) denote the equivalence class of A (B). Then M is a fuzzy metric on A/μ.

The following Lemma shows that the collection of equivalence class A/μ forms a fuzzy σ-algebra.

Lemma 3.5: Let (X A μ) be an F -measure space. Then for Ai, BiA, AiBi, i ∈ ℕ, we have

  1. 1 - Ai1 - Bi

  2. A1A2 ~ B1B2 and A1A2 ~ B1B2

  3. i=0Aii=1Bi.

Lemma 3.6: Let (X, A μ) be an F-measure space. If, for Ai, BiA, AiBi, i ∈ ℕ, AiA, BiB and then AB.

According to Lemma 3.5 and Lemma 3.6, by means of representatives of classes, we can introduce the operations of union, intersection and complementation on A/μ:i=1[Ai]=[i=1Ai];i=1[Ai][i=1Ai];1[Ai]=[1Ai], [Ai] ∈ A/μ denote the equivalence class of Ai in A. Hence, A/μ is a fuzzy σ algbra. We therefore properly define μ on A/μ by setting μ([A])=μ(A),forallAA.

The pair (A/μ, μ) is a said to be an F-measure algebra.

For convenience and simplicity, we denote members [A] of A/μ by A, and functions μ: A/μ → [0, ∞) by μ: A/μ → [0, ∞)

4 Properties of the fuzzy metric space( A, M, ⊤)

In this section, we study some properties of the fuzzy metric space ( A, M, ⊤) based on F -measure µ.

Theorem 4.1: Let (X, A, µ) be an F -measure space and M be the fuzzy metric on A defined in Theorem 3.4. Then the maps (i) (A, B) ↦ AB and (ii) (A, B) ↦ AB are uniformly continuous from A × A to A.

Now, we shall consider the completeness of the fuzzy metric space A, M, ⊤.

Lemma 4.2: Let (X, A, µ) be an F -measure space and Ai (i = 1, 2, ...) be elements of A. Then the fuzzy metric M as defined in Theorem 3.4 satisfies the following properties:

  1. M(A1, A2, t)M(A1, A1A2, t) for each t > 0;

  2. M(Ak, Ak+1, tk)M(i=1kAi, i=1k+1Ai, i=1kti), for all k ∈ ℕ, ti > 0.

Lemma 4.3: Let (xn)n∈ℕ be a Cauchy sequence in fuzzy metric space (x, M, ⊤). If there is a subsequence (xk(n)n∈ℕ of (xn)n∈ℕ that converges to x in X, then the Cauchy sequence (xn)n∈ℕ converges to x.

Theorem 4.4: The fuzzy metric space (A, M, ⊤) based on F -measure µ is complete.

5 The correspondence between fuzzy metric space and F-measure space

In the following, we give the characteristics of the nonatom of the F-measure algebra ( A, µ).

Definition 5.1: If, for two distinct elements A, B ⊂ X, there exists t, s > 0 and an element P ⊂ X, different from both A and B such that M(A, B, t + s) = M(A, P, t)⊤M(P, B, s), then fuzzy metric space (X, M, ⊤) is called convex..

Definition 5.2: Let ( A, µ) be an F-measure algebra. Then the measure µ is called nonatom if for A, B ∈ A, A ≤ B and µ(A) < µ(B), there exists P ∈ A, A ≤ P ≤ B such that µ(A) < µ(P) <(B).

Theorem 5.3: The F-measure algebra ( A, µ) is nonatom if and only if fuzzy metric space ( A, M, ⊤) is convex.

6 Conclusion

In this paper, by constructing a fuzzy metric on the fuzzy measurable sets, we studied the relations between these two. In particular, several satisfactory properties of the constructed fuzzy metric have been obtained. In addition, we investigated the charaterization of the nonatom of the fuzzy measure and the corresponding properties of constructed fuzzy metric space. The main results and methods presented in this paper generalize some well known results in [38, 39].

Acknowledgement

The authors are enormously grateful to the editors and the anonymous reviewers for their professional comments and valuable suggestions. This work is supported in part by the National Natural Science Foundation of China (No. 11371130, 61103052, 11401195), the Natural Science Foundation of Fujian Province (No. 2014H0034, 2016J01022), the projects of Education Department of Fujian Province (No. JA15280) and Li Shangda Discipline Construction Fund of Jimei University.

References

  • [1]

    Deng Z. K., Fuzzy pseudometric spaces, Journal of Mathematical Analysis and Applications, 1982, 86, 74–95

  • [2]

    Kaleva O., Seikkala S., On fuzzy metric spaces, Fuzzy Sets and Systems, 1984, 12, 215–229

  • [3]

    Kramosil I., Michálek J., Fuzzy metric and statistical metric spaces, Kybernetika, 1975, 11, 336–334

  • [4]

    Yue Y. L., Shi F. G., On fuzzy pseudo-metric spaces, Fuzzy Sets and Systems, 2010, 161, 1105–1116

  • [5]

    George A., Veeramani P., On some results in fuzzy metric spaces, Fuzzy Sets and Systems, 1994, 64, 395–399

  • [6]

    George A., Veeramani P., On some results of analysis for fuzzy metric spaces, Fuzzy Sets and Systems, 1997, 90, 365–368

  • [7]

    Gregori V., Romaguera S., Sapena A., Uniform continuity in fuzzy metric spaces, Rendiconti dell'Istituto di Matematica dell'University di Trieste Suppl., 2001, 2, 81–88

  • [8]

    Gregori V., Sapena A., On fixed point theorems in fuzzy metric spaces, Fuzzy Sets and Systems, 2002, 125, 245–253

  • [9]

    Gregori V., Morillas S., Sapena A., Examples of fuzzy metrics and applications, Fuzzy Sets and Systems, 2011, 170, 95–111

  • [10]

    Shen Y. H., Qiu D., Chen W., Fixed point theorems in fuzzy metric spaces, Applied Mathematics Letters, 2012, 25, 138–141

  • [11]

    Shen Y. H., Qiu D., Chen W., Fixed point theory for cyclic ϕ-contractions in fuzzy metric spaces, Iranian Journal of Fuzzy Systems, 2013, 10, 125–133

  • [12]

    Sugeno M., Theory of fuzzy integrals and its applications, Ph.D. Thesis, Tokyo Institute of Technology, Tokyo, 1974.

  • [13]

    Klement E. P., Fuzzy σ-algebras and fuzzy measurable functions, Fuzzy Sets and Systems, 1980, 4, 83–93

  • [14]

    Klement E. P., Characterization of finite fuzzy measures using markoff-kernels, Journal of mathematical analysis and applications, 1980, 75, 330–339

  • [15]

    Srivastava Y. K., Fuzzy probability measures and dynamical systems, Ph.D. Thesis, Allahabad University, 1993.

  • [16]

    Klir G. J., Generalized information theory, Fuzzy Sets and Systems, 1991, 40, 127–142

  • [17]

    Dubois D., Prade M., Sabbadin R., Decision-theoretic foundations of possibility theory, European Journal of Operational, 2001, 128, 459–478

  • [18]

    Dubois D., Prade M., A class of fuzzy measures based on triangular norms, International Journal of General System, 1982, 41, 59–95

  • [19]

    Dubois D., Prade M., Aggregation of decomposable measures with application to utility theory, Theory and Decision, 1996, 41, 59–95

  • [20]

    Weber S., ┴-decomposable measures and integrals for archimedean t-conorm, Journal of Mathematical Analysis and Applications, 1984, 101, 114–138

  • [21]

    Sugeno M., Murofushi T., Pseudo-additive measures and integrals, Journal of Mathematical Analysis and Applications, 1987, 122, 197–222

  • [22]

    Pap E., Pseudo-analysis, Computational Cybernetics (ICCC), 2013 IEEE 9th International Conference, IEEE, 2013, 213–218

  • [23]

    Butnariu D., Klement E. P., Triangular Norm-Based Measures and Games with Fuzzy Coalitions, Springer Netherlands, 1993.

  • [24]

    Klement E. P., construction of fuzzy σ-algebras using triangular norms, Journal of mathematical analysis and applications, 1982, 85, 543–565

  • [25]

    Klement E. P., Characterization of fuzzy measures constructed by means of triangular norms, Journal of mathematical analysis and applications, 1982, 86, 345–358

  • [26]

    Klement E. P., Weber S., Generalized measures, Fuzzy Sets and Systems, 1991, 40, 375–394

  • [27]

    Denneberg D., Non-additive measure and integral, Kluwer Academic Publishers, 1994

  • [28]

    Pap E., Null-additive set functions, Ister Science, 1995

  • [29]

    Pap E., Handbook of measure Theory: In two volumes, volume 2, Elsevier Science, Amsterdam, 2002

  • [30]

    Wang Z. Y., Klir G. J., Generalized Measure Theory, Spalgebraer Press, New York, 2009

  • [31]

    Repovš D., Savchenko A., Zarichnyi M., Fuzzy prokhorov metric on the set of probability measures, Fuzzy Sets and Systems, 2011, 175, 96–104

  • [32]

    Savchenko A., Zarichnyi M., Fuzzy ultrametrics on the set of probability measures, Topology, 2009, 48, 130–136

  • [33]

    Cao Y. Z., Sun S. X., Wang H., Chen G., A behavioral distance for fuzzy-transition systems, IEEE Transactions on Fuzzy Systems, 2013, 21, 735–747

  • [34]

    Qiu D., Zhang W. Q., Li C., On decomposable measures constructed by using stationary fuzzy pseudo-ultrametrics, International Journal of General Systems, 2013, 42, 395–404

  • [35]

    Qiu D., Zhang W. Q., Li C., Extension of a class of decomposable measures using fuzzy pseudometrics, International Journal of General Systems, 2013, 222, 33–44

  • [36]

    Xie J. L., Li Q. G., Chen S. L., Gao Y., On pseudo-metric spaces induced by σ┴-decomposable measures, Fuzzy Sets and Systems, 2016, 289, 33-42

  • [37]

    Klement E. P., Mesiar R., Pap E., Triangular norms, Kluwer Academic Publishers, Dordrecht, 2000

  • [38]

    Halmos P. R., Measure Theory, Spalgebraer-Verlag, New York, 1974

  • [39]

    Royden H. L., Real Analysis, Macmillan, NewYork, 1989

About the article

Received: 2016-05-14

Accepted: 2016-07-05

Published Online: 2016-09-20

Published in Print: 2016-01-01


Citation Information: Open Mathematics, ISSN (Online) 2391-5455, DOI: https://doi.org/10.1515/math-2016-0051. Export Citation

© 2016 Xie etal., published by De Gruyter Open. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. (CC BY-NC-ND 3.0)

Comments (0)

Please log in or register to comment.
Log in