Abstract
It is widely known that Feature Selection plays a substantial role in different fields of Artificial Intelligence. Extracting beneficial subset of features from the main feature set, not only can simplify the created system and makes it more comprehensible, but also can greatly enhance the accuracy level of final system by eliminating noises. More importantly, this issue gets more critical when the medical data sets are involved. During recent years, lots of efforts have been dedicated to develop some modern feature selection algorithm for better identification of optimum subset of features for complex nonlinear systems. In this research we investigate to propose a new hybrid Genetic Algorithm-Artificial Neural Network based feature selection method to find the best subset of features. The proposed method builds some primary systems, by means of fuzzy system modeling to provide enough samples for training an Artificial Neural Network. Then, the trained network will be used as the fitness function of GA to assess each subset of features. Afterwards, to find the optimum subset of features, the Genetic Algorithm will be initiated and its different phases will be implemented
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
