Conference Paper

Credit Rating Analysis with AFS Fuzzy Logic.

DOI: 10.1007/11539902_152 Conference: Advances in Natural Computation, First International Conference, ICNC 2005, Changsha, China, August 27-29, 2005, Proceedings, Part III
Source: DBLP

ABSTRACT In this paper, we propose a new machine learning approach based on AFS (Axiomatic Fuzzy Sets) fuzzy logic, in attempt to pro- vide a better model with interpretability. First, we will concisely present the AFS theory. Second, we will propose new membership functions for fuzzy sets and their logic operations. Third, we will design a new machine learning algorithm based on the new membership functions and their logic operations. This algorithm has two advantages. One is that it can mimic the human reasoning comprehensively and offers a far more flex- ible and effective means for the study of large-scale intelligent systems. Another is its simplicity in implementation and mathematical beauty in fuzzy theory. Finally, a credit data example is used to illustrate its effectiveness.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Many research results of AFS (Axiomatic Fuzzy Set) theory and its applications have been published and reported since Liu proposed it in (20) in 1995. In this paper, an over review of AFS theory is done by both theory analysis and illustrate examples to explain the abstract notations and theorems in order to elicit the potential applications and the further research topics. Many well-known datasets are applied to test the application algorithms and the results show that AFS fuzzy logic system oers a far more flexible and powerful framework for representing human knowledge and studying the large- scale intelligence systems in real world applications.
    International Journal of Information & Systems Sciences. 01/2006; 2(3).
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a new classifler design based on fuzzy clustering approaches via AFS theory proposed by Ren Yan et. al. (2006). Firstly, the new fuzzy classifler, which is based on the fuzzy descriptions of the classes, is proposed. The classifler has two main advantages: One is that it can mimic the human reasoning processes and ofier an interpretable classifler which is represented by some fuzzy sets with deflnitely semantic interpretations. An- other is the data types of the attributes can be various types or sub-preference relations, even descriptions of human intuitions. Finally, Iris data-set is used to illustrate accuracy and stability of the new classifler. A number of illustrative examples show that this approach ofiers a far more ∞exible and efiective means for the intelligent systems in real-world applications.
    International Journal of Information & Systems Sciences. 01/2007; 3(4).
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: As moving further into the age of machine intelligence and auto- mated decision-making, we have to deal with both the subjective imprecision of human perception-based information described in natural language and the ob- jective uncertainty of randomness universally existing in the real world. A basic limitation of standard probability theory which cannot deal with information described in natural language becomes a serious problem. With its abilities to represent natural language, the notion of AFS (Axiomatic Fuzzy Set) theory has proven useful in the clustering, classiflcations, concept representations and decision trees. In this paper, we apply AFS theory and probability theory to propose a new interpretation of the membership functions taking both fuzziness (subjective imprecision) and randomness (objective uncertainty) into account. So that uncertainty of randomness and of imprecision can be treated in a uni- fled and coherent manner under the AFS and probability framework. It opens a door to explore the deep mathematical analysis properties of fuzzy set the- ory and to a major enlargement of the role of natural languages in probability theory.
    International Journal of Information & Systems Sciences. 01/2007; 3(2).

Full-text (2 Sources)

Available from
May 29, 2014