Abstract
Fuzzy logic has gained tremendous popularity in recent years as its applications are found in areas ranging from consumer products to industrial process control and portfolio management. Along with neural networks and genetic algorithms, fuzzy logic constitutes three cornerstones of “soft computing.” Unlike the traditional or hard computing, soft computing strives to model the pervasive imprecision of the real world. Solutions derived from soft computing are generally more robust, flexible, and economical. In addition, constituent technologies of soft computing are generally complementary rather than competitive. As a result, many hybrid systems have been proposed to integrate these complementary technologies. This study reviews fuzzy logic and neural networks and illustrates how they can be integrated to provide a better solution. In an empirical test, the integrated neural fuzzy system significantly outperformed a traditional statistical model in predicting pension accounting adoption choices.