The prediction of school district division can provide a scientific basis for educational planning and development.The traditional K-means algorithm has the problem of low prediction hit rate when making school district division predictions.In view of this,hierarchical clustering was used to optimize the processing process of K-means algorithm.Later,the Euclidean dis-tance used by the traditional K-means algorithm is improved to the path planning distance,which greatly improves the prediction hit rate of the overall algorithm.Experimental results show that the designed hierarchical K-means algorithm improves the predic-tion hit rate by 4.7%on average compared with the original K-means algorithm when predicting school districts,and the hierarchi-cal K-means algorithm integrated with path planning distance improves the prediction hit rate by 15.63%on average compared with the original K-means algorithm.
关键词
K-means算法/分层聚类/路径规划距离/学区划分
Key words
K-means algorithm/hierarchical clustering/path planning distance/school district division
万方数据