摘要
近年来,针对移动通信网络的投诉量持续增加,而网络优化人员使用的投诉热点区域通常为没有明确边界范围的模糊区域.为快速识别移动通信网络的用户投诉热点区域,并圈定精准范围,基于DBSCAN聚类和凸包算法,提出一种投诉热点区域智能识别方法.利用DBSCAN算法对历史投诉数据集进行预处理,去除噪声;然后,基于Graham扫描方法求解了投诉热点区域的边界点,并绘制了投诉热点区域的边界图.投诉热点区域识别后,基于投诉热点区域网络优化模型,可以快速输出投诉热点区域解决方案.
In recent years,the number of complaints against mobile communication network continues to increase,while the hot area of complaints used by network optimization personnel is usually a fuzzy area without a clear boundary.In order to quickly identify the user complaint hotspots of mobile communication network and delineate the precise scope,an intelligent recognition method of complaint hotspots is proposed based on DBSCAN clustering and convex hull algorithm.First,the DBSCAN algorithm is used to preprocess the historical complaint data set and remove the noise.Then,the boundary points of complaint hotspot area are solved based on Graham scan method and the boundary of complaint hotspot area is sketched.Finally,based on the optimization model of complaint hotspot region,the solution of complaint hotspot region can be output quickly.
作者
孟现锋
梁松柏
徐刚
MENG Xianfeng;LIANG Songbai;XU Gang(Zhengzhou Branch Company of China Unicom,Zhengzhou 450000,China;Henan Branch Company of China Unicom,Zhengzhou 450000,China)
出处
《河南科学》
2021年第8期1211-1216,共6页
Henan Science
基金
中国联通CAPEX研发项目(X9120C1B9Z0019)。
关键词
DBSCAN算法
凸包算法
移动通信网络
投诉热点
DBSCAN algorithm
convex hull algorithm
mobile communication network
complaint hot pots
