论文部分内容阅读
提出一种基于二叉树支持向量机的超球孪生二叉树支持向量机,该算法结合了孪生支持向量机和二叉树支持向量机的优势,加快了训练速度,减少了误差累计.通过引入坐标轮换法和收缩技术,得到超球坐标轮换孪生二叉树支持向量机.实验结果表明,这两种算法具有如下优点:相比一对多支持向量机,在训练时间上具有绝对的优势,特别是在处理数据规模较大且稀疏性较强的问题时;避免了一对多支持向量机可能存在的样本不均衡性、不可分区域等缺点.
This paper proposes a new hyperblastic twinning tree support vector machine based on binary tree support vector machine which combines the advantages of twin support vector machines and binary tree support vector machines to speed up the training and reduce the error accumulation.Through the introduction of coordinate rotation and shrinkage Technology, we get the twins binary tree support vector machine with hypersphere coordinate rotation.The experimental results show that these two algorithms have the following advantages: Compared with one-to-many support vector machines, they have the absolute advantage in training time, especially when dealing with data size Large and sparsity problem; avoiding the possible sample imbalance of one-to-many support vector machines, inseparable regional shortcomings.