知识卡片
K均值交替分配与更新中心
先给k个初始中心,将点分到最近中心,再按各簇均值更新,重复到满足停止条件。它优化簇内平方距离,初始化不必须两两最远,局部收敛也不保证全局最优。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 机器学习基础
原书出处 · 1 处
- · 查看原书依据
K-means is a partition-based clustering technique that uses Euclidean distances between the points as a criterion for the formation of clusters. K-means groups n number of data points into a predetermined k number of clusters. A center is allocated to each cluster and each cluster is placed at farthest distance. The incoming data point is placed in a cluster with the closest cluster center. This
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第02章《机器学习基础》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/13_chapter02.xhtml)
- 结论依据:摘录中关于K均值交替分配与更新中心的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:K-means is a partition-based clustering technique that uses Euclidean distances between the points as a criterion for the formation of clusters. K-means groups n number of data points into a predetermined k number of clusters. A center is allocated to each cluster and each cluster is placed at farthest distance. The incoming data point is placed in a cluster with the closest cluster center. This