Performance Assessment of Voting Algorithms in Artificial Intelligence through Neural Network

Abstract

            In this paper, the techniques that employ Artificial Neural Network to obtain voting methods are better way to improve classification algorithm performance. These classification algorithms have usually been applied for complete datasets. Findings effective method for developing a sample of models has been a present study area of large scale data mining in recent years. In this paper, to categorize the instances based on the classes, which are given in our complete datasets. Technically this approach is called voting. We propose a new voting methodology, which combines the feature of standard propagation, Neighborhood based standard backpropagation and neighborhood based learning coefficient (K_NN) to train single hidden layer neural network.