Banking applications of FCM models

Miklós F. Hatwágner, Gyula Vastag, Vesa A. Niskanen, L. Kóczy

Research output: Chapter

1 Citation (Scopus)

Abstract

Fuzzy Cognitive Map (FCMs) is an appropriate tool to describe, qualitatively analyze or simulate the behavior of complex systems. FCMs are bipolar fuzzy graphs: their building blocks are the concepts and the arcs. Concepts represent the most important components of the system, the weighted arcs define the strength and direction of cause-effect relationships among them. FCMs are created by experts in several cases. Despite the best intention the models may contain subjective information even if it was created by multiple experts. An inaccurate model may lead to misleading results, therefore it should be further analyzed before usage. Our method is able to automatically modify the connection weights and to test the effect of these changes. This way the hidden behavior of the model and the most influencing concepts can be mapped. Using the results the experts may modify the original model in order to achieve their goal. In this paper the internal operation of a department of a bank is modeled by FCM. The authors show how the modification of the connection weights affect the operation of the institute. This way it is easier to understand the working of the bank, and the most threatening dangers of the system getting into an unstable (chaotic or cyclic state) can be identified and timely preparations become possible.

Original languageEnglish
Title of host publicationStudies in Computational Intelligence
PublisherSpringer Verlag
Pages61-72
Number of pages12
DOIs
Publication statusPublished - jan. 1 2019

Publication series

NameStudies in Computational Intelligence
Volume796
ISSN (Print)1860-949X

    Fingerprint

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Hatwágner, M. F., Vastag, G., Niskanen, V. A., & Kóczy, L. (2019). Banking applications of FCM models. In Studies in Computational Intelligence (pp. 61-72). (Studies in Computational Intelligence; Vol. 796). Springer Verlag. https://doi.org/10.1007/978-3-030-00485-9_7