大型呼叫中心话务量预测

Forecast on traffic of large call center

为进行呼叫中心的坐席数估计和后续的排班工作,分析了历史话务量数据的特点,总结出影响大型呼叫中心话务量的因素,并用这些影响因素的不同组合来预测话务量,通过结果的对比分析得出相对最优的话务量预测模型。在此模型的基础上分别采用BP神经网络算法和支持向量机算法(LS-SVM)对话务量进行了预测,通过分析和比较结果表明,BP神经网络比支持向量机算法更适合对大型呼叫中心话务量的预测。

To estimate the number of agents and the follow-up scheduling of a call center, the characteristics of traffic＇ s historical data are analyzed. Some factors impacting the traffic of large call center are summed up, and different combinations of these factors are used to predict the traffic, compared the results, a optimum model is obtained. Based on the model, the BP neural network algorithm and LS-SVM algorithm are used to the traffic forecast, besides a comparative analysis is made for the two methods. According to the presented simulations, it is proved that BP neural network algorithm is more efficient to the traffic forecasting for large call center than LS-SVM.