Cause-effect framework dividing method supporting high-dimensional cause-effect discovery

• Main Authors: MAI GUIZHEN, HONG YINGHAN, GUO CAI, CHEN PINGHUA, PENG SHIGUO
• Format: Patent
• Language: Chinese; English
• Published: 27.04.2018
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a cause-effect framework partitioning method supporting a high-dimensional cause-effect discovery. According to the cause-effect framework partitioning method, a problem domainis divided into small subproblems by utilizing CI tests, and partial results of the subproblems are finally merged, thereby a complete causal relationship of original data is returned. Experiments ofa real cause-effect network validate the superior extensibility and effectiveness of the method proposed by the invention. The cause-effect framework partitioning method has the beneficial effects that: the fact that a CDF framework theory is sound and perfect is proved; cause-effect partitions returned by means of the CDF is more reliable; and the CDF is a fast framework supporting the high-dimensional cause-effect discovery, the robustness of data analysis is high, and the correctness is high. 本发明公开了种支持高维度因果发现的因果框架划分方法，用CI测试将问题域划分为小的子问题，从各子问题最后合并在起的部分结果，返回了关于原始数据的完整因果关系。真实因果网络的实验验证了我们建议的卓越可扩展性和有效性。有益效果：证明了CDF框架理论健全和