基于ARMA模型的电视台收视率预测方法设计和实现

TV Ratings Prediction Method Based on ARMA

收视率是电视行业中最重要的指标之一,目前行业内预测收视率的方法均没有考虑收视率自身的长期发展趋势,忽略了预测过程中重要的记录信息,导致预测结果精度不够,充分考虑收视率数据自身长期稳定性的特点,使用ARMA模型来进行预测,使预测精度达到用户期望。ARMA模型是一类常用的随机时序模型,它是一种精度较高的时序预测方法,预测过程中充分考虑序列内在的发展趋势,对于具有长期稳定性的时间序列的预测非常有效,并已经在各个行业得到了应用,预测精度较高。收视率属于长期稳定,起伏波动不大的时间序列,引入ARMA模型进行预测必将改善传统预测方法得到结果的精度,仿真结果表明使用ARMA模型可以提高收视率预测精度。

The audience rating is one of the most important indicator of the television industry.At present most prediction stradegies do not consider the ratings of its own long-term development trends and overlook some important informations,leading to lacking of accuracy.Considering the long-term stability of the rating data,the ARMA model is used to forecast and the results can satisfy custom expectation.The objective is to improve the forecast accuracy.ARMA model is a commonly used random timing model and a high precision timing prediction method,taking full consideration in the course of the development trend of internal sequence.It's very effective for long-term stability of the predicted time series and been applied in many sectors.It can improve the accuracy of traditional prediction stradegies for TV ratings. And the simulation results show that ARMA model can improve the forecast accuracy of the ratings.