| Summary: | 碩士 === 靜宜大學 === 企業管理研究所 === 91 === This thesis is trying to construct a financial crisis warning system by using neuro fuzzy which is a hybrid technique combining the functionality of fuzzy logic and the learning ability of neural network. The empirical results show that the proposed neuro fuzzy model can have the least classification error cost among these three competitive methods, neuro fuzzy, neural network, and logistic regression, in addition to the highest hit rate. Besides, the proposed neuro fuzzy model can also have earlier and increasingly stronger signals than the others when the crisis approaches. The three-dimension graph among the variables and the knowledge base rules obtained from the neuro fuzzy also provide another perspective to tackle the financial crisis problem.
|
|---|
