基于支持矩的无误差支持向量机增量算法

Incremental Svm Algorithm Without Error Based on Support Rectangle

为了提高支持向量机的训练速度,并有效地提高预测结果的精确度,提出了支持向量机支持矩的概念以及相应的支持向量机算法,仿真实验证明,该算法相比传统算法在明显提高了训练速度的情况下,保留了预测精度而没有产生新的误差,获得了很好的预测效果。

To increase the training speed of support vector machine,and efficiently improve the accuracy of the prediction result,support matrix of support vector machine is proposed.result from the experiment shows that this algorithm obviously increase the training speed compared with the traditional algorithm,under such conditions,retain the prediction accuracy without new er ror found,get a good performance for prediction.