Abstract
Neural network is one of the pattern classification techniques, which has been widely used in many fields application. The multilayer perceptron in a training process has an accuracy impact. The selection of features is another factor that influences classification accuracy. The objective of this research is to optimize simultaneously parameters and subset of functionalities in order to increase multilayer perceptron accuracy. A genetic algorithm approach is presented in order to characterize selection and optimization parameters that serve as input to the multilayer perceptron classifier. The proposed algorithm can scan irrelevant information in PIMA database and obtain better accuracy. The results of this experience show that the proposed algorithm can provide a significant performance gain in terms of accuracy and training speed for the multilayer perceptron classification.
Published Version
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