论文部分内容阅读
为了科学准确地预测近几年因特网访问人数,提出了应用灰色马尔可夫Verhulst模型进行预测的方法。首先,利用历史数据建立灰色Verhulst模型,通过确定系数可获得因特网访问人数的时间响应序列的表达式,从而可获得未来年份因特网访问人数的发展序列值。然后,结合马尔可夫链过程将序列状态划分为三类,通过确定状态转移矩阵可获得序列处于各状态的概率值及与各状态对应的预测中值,最终求得各序列的修正值。最后,通过2006/12~2012/6期间我国互联网上网人数的历史数据,预测了最近四个统计时段的访问人数。实例表明,该模型预测结果的误差更小、精度更高,还能提供预测结果的波动范围及出现概率,能够为网络建设及管理提供决策依据。
In order to predict the number of Internet visitors in recent years scientifically and accurately, a method based on Gray Markov Verhulst model is proposed. First, the gray Verhulst model is established by using historical data, and the expression of the time response sequence of the number of Internet visitors can be obtained by determining the coefficient so as to obtain the development sequence value of the number of Internet visitors in the next year. Then, the state of the system is divided into three categories according to the Markov chain process. By determining the state transition matrix, the probability values of the sequences in each state and the predicted median corresponding to each state can be obtained. Finally, the correction value of each sequence is obtained. Finally, based on the historical data of Internet users in China from 2006/12 to 2012/6, the number of visitors in the last four statistical periods is predicted. The example shows that the model has less error and higher accuracy, and can provide the fluctuation range and probability of prediction, which can provide decision basis for network construction and management.