Informatica

Fuzzy relations have been widely applied in decision making process. However, the application process requires people to have a high level of ability to compute and infer information. As people usually have limited ability of computing and inferring, the fuzzy relation needs to be adapted to fit the abilities of people. The bounded rationality theory holding the view that people have limited rationality in terms of computing and inferring meets such a requirement, so we try to combine the fuzzy relation with the bounded rationality theory in this study. To do this, first of all, we investigate four properties of fuzzy relations (i.e. reflexivity, symmetry, transitivity and reciprocity) within the bounded rationality context and find that these properties are not compatible with the bounded rationality theory. Afterwards, we study a new property called the bounded rational reciprocity of fuzzy relations, to make it possible to combine a fuzzy relation with the bounded rationality theory. Based on the bounded rational reciprocity, the bounded rational reciprocal preference relation is then introduced. A rationality visualization technique is proposed to intuitively display the rationality of experts. Finally, a bounded rationality net-flow-based ranking method is presented to solve real decision-making problems with bounded rational reciprocal preference relations, and a numerical example with comparative analysis is given to demonstrate the advantages of the proposed methods.

In Computer Vision and Pattern Recognition, surveillance-video crowded scenes have been analysed according to their structure, where the detection of distinguishable people groups is an essential step. In this paper, we are interested in detecting F-Formations (i.e. free standing conversational groups) on video, which are formed by people social relations. We proposed a new method based on fuzzy relations, where a new social representation for computing relation between individuals, fusion for search consensus in multiple frame and clustering are introduced. Finally, our proposal was tested in a real-world dataset, improving the already reported scores from literature.