图 书 馆
将图形拟合到矢量数据Fitting a Graph to Vector Data |
|
| 课程网址: | http://videolectures.net/icml09_daitch_fgv/ |
| 主讲教师: | Samuel I. Daitch |
| 开课单位: | 耶鲁大学 |
| 开课时间: | 2009-08-26 |
| 课程语种: | 英语 |
| 中文简介: |  我们引入了一种衡量组合图如何处理向量集合的方法。这种度量下的最优图可以通过求解凸二次规划来计算，并且具有许多有趣的特性。对于 d 维空间中的向量，图的平均度最多为 2(d 1)，而对于 2 维空间中的向量，它们总是平面的。我们为许多标准数据集计算了这些图，并表明它们可用于获得分类、回归和聚类问题的良好解决方案。 |
| 课程简介: | We introduce a measure of how well a combinatorial graph ts a collection of vectors. The optimal graphs under this measure may be computed by solving convex quadratic programs and have many interesting properties. For vectors in d dimensional space, the graphs always have average degree at most 2(d+1), and for vectors in 2 dimensions they are always planar. We compute these graphs for many standard data sets and show that they can be used to obtain good solutions to classifi cation, regression and clustering problems. |
| 关 键 词: | 衡量组合图; 凸二次规划; d 维空间 |
| 课程来源: | 视频讲座网 |
| 数据采集: | 2021-08-11：nkq |
| 最后编审: | 2021-08-28：nkq |
| 阅读次数: | 51 |