外卖行业蓬勃发展，人们对于外卖服务质量的要求也不断提升。因外卖员接单区域基本固定，合理划分外卖员负责区域并分配每个区域外卖员人数成为提升效率的关键。本项目基于查询算法模型，分析上海市2017年某时段的外卖数据，试图得到一个对于外卖接单区域的较为合理的划分标准并给出该划分。K-Means是一种常见的划分聚类算法，是在集中式系统框架无法对海量数据进行处理分析的基础上提出的。然而对于有权重的二维点集无法使用K-Means聚类算法，因此研究一种改进的Weighted K-Means算法显得尤为必要。本项目定义带权质心和带权距离，提出了新的Weighted K-Means算法，并使用改进前后的两种方法处理上海市外卖接单信息，给出合理可行的外卖员接单区域划分。对比两种方法的结果，改进的Weighted K-Means不仅方法可行，区域划分表现也更优秀。与此同时，使用该方法对外卖接单区域进行新的划分，有助于优化现有外卖模式、提升外卖效率以及顾客满意度。
The take-away industry is booming and
people's demands for the quality of take-away services are constantly
improving. Since the take-out area of the take-out is basically fixed, it is
the key to improve efficiency by reasonably dividing the area of the take-out
of the seller and assigning the number of the sellers in each area. Based on
the query algorithm model, this project analyzes the take-out data of Shanghai
during a certain period of time in 2017, trying to obtain a more reasonable
division standard for the take-out order area and give the division. K-Means is
a common partitioning clustering algorithm based on the inability of
centralized system frameworks to process and analyze massive amounts of data.
However, the K-Means clustering algorithm cannot be used for the
two-dimensional point set with weight information. Therefore, it is especially
necessary to study an improved Weighted K-Means algorithm. This project defines
the weighted centroid and weighted distance, proposes a new Weighted K-Means
algorithm, and uses the two methods before and after the improvement to deal
with the take-out information of Shanghai take-out orders, and gives a reasonable
and feasible division of the take-out area of the take-out. Comparing the
results of the two methods, the improved Weighted K-Means is not only feasible,
but also better in regional division. At the same time, using this method to
make a new division of the take-out order area helps to optimize the existing
take-away model, improve the take-out efficiency and customer satisfaction.
