地理加权的K-Modes算法在城市餐饮空间分析中的应用

Application of Geographically Weighted K-Modes Algorithm in Urban Catering Space Analysis

通过Python获取大连市餐饮空间数据,在SPSS中采用K-Modes聚类分析算法进行了第一次基础聚类,以经纬度作为地理权重进行二次聚类,在Arcgis中分析了大连市线上餐饮业空间消费规律。通过划定消费区间,在固定区间内实现附加地理权重的K-Modes聚类,将线上餐饮消费空间差异程度进一步放大。研究发现:附带地理权重的K-Modes聚类算法能更精准地体现餐饮消费规律地理空间上的差异性。

The data of Dalian's catering space is obtained through Python.The K-Modes clustering analysis algorithm in SPSS is used to perform the first basic clustering.Latitude and longitude is used as the geographic weight to perform secondary clustering.The law of industry space consumption is analyzed by ArcGIS in Dalian online catering industry.By delimiting the consumption interval,K-Modes clustering with additional geographical weights is realized within a fixed interval,which further amplifies the spatial differences of online catering consumption.The study found that the K-Modes clustering algorithm with geographic weights can more accurately reflect the geographic and spatial differences of catering consumption patterns.