Showing papers presented at "Granular Computing in 2023"

Multi-attribute decision-making for electronic waste recycling using interval-valued Fermatean fuzzy Hamacher aggregation operators

Abstract: The utilization of electrical and electronics equipments in waste recycling has become a paramount for various countries. The waste electrical and electronics equipment (WEEE) recyclers own a crucial position in the environmental growth of a country as they help to minimize the carbon emissions during the recycling of WEEE in the most eco-friendly way. Therefore, the selection and assessment of an appropriate WEEE recycling partner has become a most important part of DM (decision-making) applications. The collusion of numerous quantitative and qualitative factors makes the recycling partner selection problem, a multifaced and significant decision for the managerial experts. The main objective of this work is to propose MADM (multi-attribute decision-making) techniques to evaluate the WEEE recycling partners under interval-valued Fermatean fuzzy (IVFF) information. In this regard, certain Hamacher AOs (aggregation operators) are proposed to develop the required DM method. These AOs include Hamacher weighted averaging, ordered weighted averaging, weighted geometric, ordered weighted geometric, generalized Einstein weighted averaging, generalized Einstein ordered weighted averaging, generalized Einstein weighted geometric, etc. Then, these averaging operators are utilized to come up with a MADM techniques under IVFF environment. Furthermore, the constructed technique is applied to a case study in China to incorporate with the e-waste recycling partner selection problem. Moreover, a brief comparison of the proposed with is presented with various existing techniques to manifest the productivity and coherence of the proposed model. Finally, the accuracy and consistency of results shows that the proposed technique is fully compatible and applicable to handle any MADM problem.

A supervised method to enhance distance-based neural network clustering performance by discovering perfect representative neurons

Abstract: Abstract Distance-based neural network clustering requires the intrinsic assumption that a particular neuron in the network represents a cluster centroid. However, not all these neurons can perfectly represent the training data; these neurons can only represent part of the training samples. This paper proposes an effective training data splitting method (TDSM) to find perfect representative neurons and improve the clustering results in a distance-based neutral network without changing the original network’s internal algorithm or the training data quality. The method allows a network with N neurons to be enlarged to a new network with $$m\times N$$ m × N neurons. These neurons represent m subnetworks, and each subnetwork perfectly represents a part of the training set, where the clustering qualification indicators (the purity, normalized mutual information, and adjusted rand index measures) all equal 1. The results are statistically validated with a t test, and we demonstrate that the TDSM performs better than the original clustering paradigm on some real datasets.

Improved rough approximations based on variable J-containment neighborhoods

Abstract: Abstract Classic generalized rough set model in neighborhood systems provides a more general framework for depicting approximations, while it may meet the non-reflexive situations. Some scholars put forward different neighborhoods, such as adhesion neighborhoods (briefly, $$P_{j}$$ P j -neighborhoods), containment neighborhoods (briefly, $$C_{j}$$ C j -neighborhoods), and $$E_{j}$$ E j -neighborhoods. However, not all of them are reflexive. Moreover, the granularity of $$P_{j}$$ P j -neighborhoods and $$C_{j}$$ C j -neighborhoods are too fine, and that of $$E_{j}$$ E j -neighborhoods too coarse. To solve the problem, we aim to design a novel construction approach of neighborhoods, called variable j -containment neighborhoods (briefly, $$V_{j}^{\beta }$$ V j β -neighborhoods), which satisfies the reflexivity and the granularity so flexible that the neighborhood space can adjust the granularity to meet the needs of problems. We generalize three kinds of rough approximations in $$V_{j}^{\beta }$$ V j β -neighborhood spaces and discuss their properties. What’s more, we analyze the topology structures relying on $$V_{j}^{\beta }$$ V j β -neighborhood spaces and compare our proposed approach with the existing approaches. By selecting the appropriate parameter $$\beta$$ β , our neighborhood system is more flexible in adjusting the granularity to fit problem requirements. And illustrative examples demonstrate the advantages of the proposed rough set model to attribute reduction in incomplete information systems.