【摘要】 分析了现有的三种预测用户下一时刻位置的方法,即概率模型、Markov模型、状态模型。在此基础上提出了一种基于普适计算上下文信息的学习模型,用logistic回归解决分类问题。该模型可以很好很灵活地预测用户的下一位置。更多还原

【Abstract】 This paper analyzed three methods for predicting next location: probability model、Markov model and state model.Proposed a learning-based model based on context of ubiquitous computing environment well.Logistic regression was using to predict labels.The learning-based model could predict next location well and freely.更多还原