摘要
提出一种基于Markov的变长预测树模型,然后根据Markov算法模型化用户的访问行为,生成一棵变长的预测树,最大程度地将用户访问路径固化在预测树上。它最大的特点是减少预测次数,提高模型的预测效率和预测精度。通过实验数据说明了所提出算法的正确性和有效性。
In this paper, a novel approach of Markov model based on predicts tree is proposed. Using predicts tree we can reduce predict times, improve predict efficiency and precision. At last, we validate the correction and efficiency of the algorithm through experiment.
出处
《科技广场》
2009年第1期72-74,共3页
Science Mosaic