基于改进的Markov模型预测准确度研究

Research on Prediction Accuracy Based on Modified Markov Model

马尔可夫(Markov)模型方法主要解决的问题是从当前状态预测下一个状态,因此适合作为一个预测模型来挖掘用户的页面访问模式。此文在总结以往Markov模型方法基础上,从页面访问时间和用户个体特征两个角度对模型方法进一步改进,以提高模型预测的准确度。最后将改进的Markov方法与协同过滤方法进行了比较,指出其在计算开销和预测准确度上的优势。

The method of Markov model resolved the problem of forecasting the next state from the current state.Thus,it's suitable for being as the forecast model to find the page visiting modes of users.The paper improves the method of model from page browsing time and user's individual characteristics on the basis of summing-up the previous method of Markov model.