Computer Engineering and Applications ›› 2016, Vol. 52 ›› Issue (12): 90-94.

Previous Articles     Next Articles

Block regression traffic prediction model for call center

ZHANG Huyin, HU Ruiyun, HE Zheng   

  1. Computer School, Wuhan University, Wuhan 430072, China
  • Online:2016-06-15 Published:2016-06-14

呼叫中心分块回归话务量预测

张沪寅,胡瑞芸,何  政   

  1. 武汉大学 计算机学院,武汉 430072

Abstract: In order to obtain the prospective traffic data, solve the seats arrangement problem of call center, realize the rational allocation of human resources, block regression traffic prediction model, based on support vector machine and K-nearest neighbor algorithm is proposed(SKBR), after analyzing the characteristics of historical traffic data. According to the date type, traffic can be divided into weekday traffic, weekend traffic and holiday traffic, and different model is used to predict the corresponding traffic. Taking the traffic of a province electric power call center for example, experiments are carried on the MATLAB platform. Results show that compared with the SVM model and improved SVM model for its method of searching parameters, SKBR model has improved the prediction accuracy.

Key words: traffic, prediction, support vector machine, nearest neighbor algorithm, prediction accuracy

摘要: 为获得前瞻性话务量数据,解决呼叫中心坐席安排的问题,实现人力资源合理配置,分析历史话务量特性,提出了基于支持向量机和[K]近邻算法的分块回归(SKBR)话务量预测模型。将话务量按日期类型分为工作日话务量、周末话务量以及节假日话务量,采用不同的模型预测相应的话务量。以某省电力呼叫中心话务量为例,在Matlab平台上进行实验。结果证明,相比SVM模型和改进寻参方法的SVM模型,SKBR模型在预测准确性上有所提升。

关键词: 话务量, 预测, 支持向量机, 近邻算法, 预测准确性