A pre-clustering technique for optimizing subclass discriminant analysis
Pattern Recognition Letters
| Hi-index | 0.00 |
As to Markov cipher, its transition probabilitymatrix is a doubly stochastic one. The eigenvalue ofthe matrix with maximum magnitude less than oneplays an important role in designing Markov cipher.This paper provides a parallel algorithm forcomputing the eigenvalue of the doubly stochasticmatrix A of size 65535脳65535, which comes from aMarkov cipher shrunken model with both 16 bitsplaintext and ciphertext, an analysis on thecomplexity of the parallel algorithm is alsoconsidered.