support vector machine is originally designed for binary classification. How to effectively extend it for multi-category classification is still an on-going research issue. This paper presents a general overview of existing representative methods for multi-category support vector machines. The processes of making decisions on the decision directed acyclic graph support vector machines were random. For this reason this paper inducts an internal-class degree of dispersion. An external-class separate measure is defined based on the distribution of the training samples to form the classes’ separating sequences. An improved algorithm having greater classification distance for decision directed acyclic graph support vector machines is proposed. The experimental results show that it has higher multi-class classification accuracy than the original decision directed acyclic graph multi-class support vector machines.