Entropy Partition Method and Its Verification for Medical Data
Jing Chen, Guangcheng Xi, Jianxin Chen · 2007
Based on the entropy partition method we proposed before, we carry out the verification process according to the lowest requirement for ideal partition result in this paper. A new definition of correlative measure among multi-variables and multi-subsystems is proposed and one full proof for its principal property "superadditivity and uniqueness" is presented. We give the full verification process to the lowest requirement for ideal partition result, thus improve this method to more dependable and more available to the complex, nonlinear, large-scale data.