algorithm - Optimal solution for clustering of rectangles
问题描述
I am looking for some approach (algorithm to be very specific) here.
Problem:There are N rectangles (r1, r2,.. rn) scattered in X-Y plane. Need to find optimal solution to cluster these rectangles with bigger bounded polygons.
Condition for clustering:
- Results should have maximum number of rectangles covered in the polygon.
- Total count of bounded polygon should be minimum possible and maximum K.
- Each bounded polygon must have at least 70% area filled with given rectangles.
- All rectangles need not to be bounded.
constraint:
- 1 millions <= n <= billions
- K = 50000
Problem can be thought of as to identify islands (max 50k island) having higher density of rectangles (70% in each island). We can off course exclude certain rectangles. But the idea is to find optimal solution and there is no single best solution.
I was trying to use K-means clustering but it doesn't fit in my case as in my problem solution can lie within 1-K values instead of K values. May be it requires all together different dimension of thinking. Hope I am clear!!
解决方案
推荐阅读
- python - 如何隐藏轴线但在 Altair 的图表中显示刻度线,同时积极使用“轴”参数?
- python - 如何将一个端口从 Elastic Beanstalk 映射到我的 Flask 应用程序的 Docker 容器?
- swift - 应用程序发送到后台并返回到前台时的IOS13音频问题
- c# - 在类库中传递和维护配置
- facebook - 如何测试使用 Facebook Groups API 的应用程序?
- c# - 与存储过程一起使用时的 FromSql 方法无法在 EF Core 3.1 中组合
- c++ - 为什么 std::isdigit 接受 int?
- r - 在小鼠/多重插补之后,如何合并 RF 和 SVL?(GLM 有效)
- eclipse - Apache wicket Tomcat 问题:java.lang.ClassNotFoundException: org.apache.wicket.protocol.http.WicketFilter
- python - pyqt5 中带有自定义小部件的拖放功能,创建内容不可见的小部件