摘要
本文提出面向多类问题的基于Chernoff上界的特征选择.推导出正态分布条件下,满足上界之和最小时的非线性矩阵方程及其迭代算法,首次获得变换矩阵的精确解.通过分析和实例可见基于Chernoff上界的特征选择是最好的特征选择.
A feature selection for minimize the sum of Chernoff upper bound of error probability of every two class pair in subspace is presented. The key of Chernoff bound feature selection is to change the problem of minimizing the criterion to a problem of solving nonlinear matrix equation with a recursive algorithm. The theoretical analysis and experimental results show that the performance of proposed algorithm is superior to the performance of any previous one.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1996年第1期26-30,共5页
Pattern Recognition and Artificial Intelligence
关键词
Chernoff上界
非线性矩阵方程
特征选择
Chermoff Bound, Upper Bound of Error Probability, Nonlinear Matrix Equation, Matrix Recursive Algorithm, Transformation Matrix.
