Cline分块矩阵的极小范数广义逆

The Minimum Norm Inverses of a Cline Partitioned Matrix

矩阵广义逆的递推算法在统计推断、模式识别以及解析动力学等领域有着广泛的应用背景。本文利用范数极小化方法给出了计算Cline分块矩阵的极小范数广义逆及逆的两个一般递推表示式,推广了已有的结果。

The recursive algorithms for computing the generalized inverses of a matrix play an important role in statistics inference, system identification, analytical dynamics and etc. In this paper, we present two more general recursive relations for computing the minimum norm generalized inverses and the inverses of the Cline partitioned matrix by utilizing the minimum-norm least square method which generalizes the existed results.