The Upper Preferred Multiple Directed Acyclic Graph Support Vector Machines for Classification

The current classification algorithms have weak fault-tolerance. In order to solve the problem, a multiple support vector machines method, called Upper preferred Multiple Directed Acyclic Graph Support Vector Machines (UMDAG-SVMs), is proposed. Firstly, we present least squares projection twin support vector machine (LSPTSVM) with confidence-degree for generating binary classifiers. It uses the idea that “when the confidence-degree outputted from the node in the directed graph, is below the threshold, the decision-making process will go on along with the two branches of the node at the same time.”, which strengthens the algorithm’s fault-tolerance. In order to select the parameters of the algorithm, we use genetic algorithm to select these parameters. Secondly, according to the minimal hypersphere distance, and the known principle “the upper-level classifiers bring up better performance of classification in DAG-SVMs ”, we present a new classification algorithm, called UMDAG-SVMs. This algorithm has two advantages of strong fault-tolerance and high classification accuracy. Finally, we make the experiments to test the performance of the algorithm. Experimental results in public datasets show that our UMDAG-SVMs has comparable classification accuracy to that other algorithms but with remarkable less computation.