Constructing overlap and grouping functions on complete lattices by means of complete homomorphisms
Introduction
Originated from problems in image processing and classification, overlap and grouping functions are two kinds of not necessarily associative fuzzy logical operators. The concepts of overlap functions [9], [10] and grouping functions [11] on the unit interval were introduced by Bustince et al. in 2009–2010 and 2012, respectively. As already stated, they are specifically important in many problems related to the fusion of information, such as image processing [10], [34], decision making [11], [29], classification [25], [26], [28], [36], [37], [38], [43], fuzzy community detection problems [33] and other fields [31], [41].
From the theoretic point of view, one of the main topics is directed towards the characterization of overlap and grouping functions on the unit interval with crucial properties, such as migrativity [8], [46], [48], [64], Archimedean property [18], modularity condition [57], homogeneity [49], relevant distributive equations [35], [47], [51], [62], [63] and so forth [5], [50]. Apart from this, interval overlap and grouping functions [2], [3], [4], [12], [44] is a hot issue as well. In addition, fuzzy implications derived from overlap and grouping functions such as -implications [24], [55], -implications [19], QL-implications [21] and other fuzzy implications [12], [23], [27], [56] have been investigated.
Researchers have attached importance to aggregation functions [6], which perform the operation of combining and merging several values into a single one. In recent years, due to the close connection between fuzzy set theory and order theory, special attention has been given to the study of aggregation functions on more general structures (e.g., bounded posets and lattices). Associative aggregation functions on bounded lattices, such as t-norms and t-conorms [16], [17], [54], [61], uninorms and nullnorms [1], [13], [14], [15], [40], semi-t-operators [30], [53], uni-nullnorms and null-uninorms [58], [60] have been studied in depth, including construction methods of them. For instance, t-norms and uninorms are often constructed by unary functions called generators. Quite recently, Sun and Liu [54] extended the additive generator theorem of t-norms to the lattices.
Different from associative aggregation functions, overlap and grouping functions may be constructed by a pair of generators [20], [22], [45]. For example, in the interval-valued context, Qiao and Hu [44] and Bedregal et al. [4] presented almost simultaneously the concept of interval-valued overlap functions. The first work also introduced the concept of interval additive generators pair.
As for overlap and grouping functions on posets, Paiva et al. introduced the notions of lattice-valued overlap and quasi-overlap functions in [42] and Wang [59] proposed constructions of overlap functions on bounded lattices. Besides, Qiao [52] gave the concrete forms of the continuity of overlap and grouping functions on complete lattices, where the continuity was replaced by meet-preserving and join-preserving. He gave two construction methods of them and extended some properties, such as -migrativity and -homogeneity, where α belongs to the complete lattice and B and C are two binary operators on the complete lattice.
As we have stated, overlap and grouping functions have experienced a rapid development and play crucial parts in many application areas. For one thing, overlap and grouping functions have good qualities of mathematical basis with its continuity. They are a specific class of binary aggregation functions which are different from the common binary aggregation function t-norms and t-conorms for their boundary conditions and removal of associativity. For another, fuzzy logics are usually based on a bounded lattice truth values. For instance, being special complete lattices, finite chains are often used as input value sets in some real problems, such as image processing and uncertainty in expert systems. Thus, it deserves our further attention to study overlap and grouping functions on complete lattices.
All up, on the basis of work in [52], in order to explore their potentialities in different scenarios, we try to develop deeper study of overlap and grouping functions on complete lattices, including construction methods and their properties. To be more specific, we extend overlap functions on complete lattices from given ones based on the simultaneous use of two complete lattice homomorphisms, named O-generator triples and investigate some properties for the overlap functions obtained by such generator triples. Dually, we propose a second construction method of overlap functions from a given grouping function based on the use of complete lattice homomorphisms. An analogous discussion goes for grouping functions on complete lattices.
The rest of this paper is structured as follows. In Section 2, we recall some basic notions needed in the sequel. In Section 3, we present the construction method of overlap functions by means of complete homomorphisms between complete lattices and some related results. In Section 4, we discuss some essential properties of this kind of overlap functions on complete lattices. In Section 5, the notion of overlap functions constructing by dual complete lattice homomorphisms is proposed. In Section 6, we tackle grouping functions generated by such operators on complete lattices in a similar way. Finally, our work is summarized.
Section snippets
Preliminaries
In this section, we briefly review several basic definitions and properties on complete lattices and overlap and grouping functions on complete lattices which will be used in the sequel. In the following, the symbol Λ always denotes a nonempty index set.
The O-generated overlap functions on complete lattices
In this section, firstly, we introduce the concept of O-generator triple of an overlap function on complete lattices, which gives a way to construct overlap functions. And then, we discuss under which conditions the two complete homomorphisms and the overlap function become the generator triple of an overlap function and show some related results. Finally, we give another way to obtain an overlap function by complete -endomorphism.
We always suppose that L and M are complete lattices in the
Some properties on O-generated overlap functions
In this section, we discuss some properties on overlap functions generated by overlap functions and complete homomorphisms, i.e., their O-generator triples, such as the -migrativity, the -homogeneity, the idempotent and cancellation law.
The G-generated overlap functions on complete lattices
In this section, we discuss overlap functions on complete lattices generated by dual complete homomorphisms and a given grouping function in a similar way as that of Section 3.
Definition 5.1 Let and be two dual complete homomorphisms between L and M and G a grouping function on M. If the two-place function given, for all , by is an overlap function on L, then is called a G-generator triple of the overlap function and is said to be
Generators of grouping functions on complete lattices
In this section, we propose construction methods of grouping functions on complete lattices by means of dual complete homomorphisms. Several related results are also showed. With the help of duality of overlap and grouping functions, proof of some theorems and propositions in this section can be omitted.
Concluding remarks
In this paper, we mainly proposed the notions of O- and G-generator triples of overlap and grouping functions, with a given overlap or grouping function and two (dual) complete homomorphisms, as a new construction method of overlap and grouping functions on complete lattices. Besides, we provided another way to extend a given overlap function or grouping function by means of complete -endomorphisms. We also analyzed some properties on overlap and grouping functions generated by their O
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors would like to express their sincere thanks to the Editors and anonymous reviewers for their most valuable comments and suggestions in improving this paper greatly. The work described in this article was supported by grants from the National Natural Science Foundation of China (grant Nos. 11971365 and 11571010).
References (64)
- et al.
General interval-valued overlap functions and interval-valued overlap indices
Inf. Sci.
(2020) - et al.
Generalized interval-valued OWA operators with interval weights derived from interval-valued overlap functions
Int. J. Approx. Reason.
(2017) - et al.
New results on overlap and grouping functions
Inf. Sci.
(2013) - et al.
A generalization of the migrativity property of aggregation functions
Inf. Sci.
(2012) - et al.
Overlap functions
Nonlinear Anal.
(2010) - et al.
On interval ()-implications and ()-implications derived from interval overlap and grouping functions
Int. J. Approx. Reason.
(2018) New methods to construct uninorms on bounded lattices
Int. J. Approx. Reason.
(2019)- et al.
A new structure for uninorms on bounded lattices
Fuzzy Sets Syst.
(2020) - et al.
New constructions of uninorms on bounded lattices
Int. J. Approx. Reason.
(2019) - et al.
Triangular norms on product lattices
Fuzzy Sets Syst.
(1999)
Ordinal sums of the main classes of fuzzy negations and the natural negations of t-norms, t-conorms and fuzzy implications
Int. J. Approx. Reason.
Archimedean overlap function: the ordinal sum and the cancellation, idempotency and limiting properties
Fuzzy Sets Syst.
QL-operations and QL-implication functions constructed from tuples and the generation of fuzzy subsethood and entropy measures
Int. J. Approx. Reason.
The law of O-conditionality for fuzzy implications constructed from overlap and grouping functions
Int. J. Approx. Reason.
On -implications derived from grouping functions
Inf. Sci.
The state-of-art of the generalizations of the Choquet integral: from aggregation and pre-aggregation to ordered directionally monotone functions
Inf. Fusion
Generalized -integrals: from Choquet-like aggregation to ordered directionally monotone functions
Fuzzy Sets Syst.
Consensus via penalty functions for decision making in ensembles in fuzzy rule-based classification systems
Appl. Soft Comput.
Semi-t-operators on bounded lattices
Inf. Sci.
Forest fire detection: a fuzzy system approach based on overlap indices
Appl. Soft Comput.
A new modularity measure for Fuzzy Community detection problems based on overlap and grouping functions
Int. J. Approx. Reason.
Some properties of overlap and grouping functions and their application to image thresholding
Fuzzy Sets Syst.
CC-integrals: Choquet-like Copula-based aggregation functions and its application in fuzzy rule-based classification systems
Knowl.-Based Syst.
Wavelet-fuzzy power quality diagnosis system with inference method based on overlap functions: case study in an AC microgrid
Eng. Appl. Artif. Intell.
Lattice-valued overlap and quasi-overlap functions
Inf. Sci.
Capacities and overlap indexes with an application in fuzzy rule-based classification systems
Fuzzy Sets Syst.
On interval additive generators of interval overlap functions and interval grouping functions
Fuzzy Sets Syst.
On multiplicative generators of overlap and grouping functions
Fuzzy Sets Syst.
On the migrativity of uninorms and nullnorms over overlap and grouping functions
Fuzzy Sets Syst.
The distributive laws of fuzzy implications over overlap and grouping functions
Inf. Sci.
On generalized migrativity property for overlap functions
Fuzzy Sets Syst.
On homogeneous, quasi-homogeneous and pseudo-homogeneous overlap and grouping functions
Fuzzy Sets Syst.
Cited by (15)
Constructing overlap functions on bounded posets via multiplicative generators
2024, International Journal of Approximate ReasoningNew results on directionally monotone quasi-grouping functions and their applications in classification problems
2023, Expert Systems with ApplicationsExtension constructions of quasi-overlap functions and their derivative concepts on function spaces
2023, International Journal of Approximate ReasoningCitation Excerpt :In 2021, Paiva et al. [42] discussed basic properties and constructions of lattice-valued overlap and quasi-overlap functions on bounded lattices. Wang and Hu [53,54] showed the constructions of overlap functions on complete lattices using complete homomorphisms and ordinal sum. Paiva et al. [41] introduced the concept of residuated implications related to quasi-overlap functions on lattices and discussed some related properties.
Constructing overlap functions via multiplicative generators on complete lattices
2022, International Journal of Approximate ReasoningCitation Excerpt :This is because we restrict the domain and range of overlap function in generator triples. Besides, we also provide examples to show that the conditions of the work by Wang and Hu in [45] can be weakened. The organization of this paper is as follows.
Minimizing a linear objective function under a max-overlap function fuzzy relational equation constraint
2022, Fuzzy Sets and SystemsCitation Excerpt :In 2019 Dimuro et al. [19] discussed the law of O-conditionality for all kinds of implications derived from overlap function. In 2021, Wang and Hu [64] obtained the overlap function constructed on complete lattices. Our research is motivated by the following statements.
Discrete overlap functions: Basic properties and constructions
2022, International Journal of Approximate ReasoningCitation Excerpt :Asmus et al. [2] introduced n-dimensional admissibly ordered interval-valued overlap functions and analyzed the influence of the combination of admissible orders and n-dimensional admissibly ordered interval-valued overlap functions in interval-valued fuzzy-rule-based classification systems. Wang and Hu [71,72] gave the constructions of overlap functions on complete lattices using complete homomorphisms and ordinal sum. Asmus et al. [3] developed a theoretical approach to aid the analysis of bivariate interval-valued operations with respect to the width of the operated intervals in order to control the uncertainty propagation, with special attention to interval-valued overlap functions.