Abstract: While most existing methods of ASVM are focus on the samples which are close to the current separating hyperplane，and it ignores some SV samples which are far form the separating hyperplane，also it doesn’t consider on if the current separating hyperplane is close to the optimal one.In order to make up for these shortages，this paper presents a new classification method of ASVM based on probability.And it not only presents a new method of probability，but also measures the degree of closeness of the current separating hyperplane to the actual separating hyperplane by a confidence factor.Experimental results verify the improvement of the proposed method both in term of classification precision and computation.

Key words: Support Vector Machine（SVM）, active learning, Active Support Vector Machine（ASVM）, confidence factor

摘要： 针对大多数主动学习支持向量机（ASVM）的主动学习策略只注重考察超平面附近的样本，忽略了有些距离超平面远但是支持向量的样本，而且没有考虑当前超平面是否接近实际的超平面。提出一种基于概率的主动支持向量机算法，采用一个置信因子来衡量当前的超平面接近实际的超平面的程度。实验结果都验证了该算法在分类精度与计算量方面都有了较大改进。

关键词: 支持向量机, 主动学习, 主动支持向量机, 置信因子