融合主动学习的改进贝叶斯半监督分类算法研究

Study on Improved Bayesian Algorithm Semi-supervised Classification Integration of Active Leaning

摘要:

半监督学习是人工智能领域一个重要的研究内容;在半监督学习中,如何有效利用未标记样本来提高分类器的泛化性能,是机器学习研究的热点和难点;主动学习可解决未标记样本有效利用的问题,将主动学习引入到半监督分类中,并改进贝叶斯算法,提出了一种基于改进贝叶斯算法的主动学习与半监督学习结合算法;实验结果表明,该方法取得了较好的分类效果。

Abstract:

As for semi-supervised learning, it is an important field of artificial intelligence research, how to make full use of unlabeled data to improve the generalization performance of classifier is a hot and difficult issue in machine learning. Active learning can effectively utilize unlabeled data, and it is introduced to semi- supervised classification, and improved Bayesian algorithm. Combination of active learning and semi-supervised learning based on improved Bayesian algorithm was proposed in this paper. Experimental results show that the algorithm is effective and feasible.