Synergetic computers and cognition: a top-down approach to neural nets
Synergetic computers and cognition: a top-down approach to neural nets
IEEE Transactions on Information Technology in Biomedicine
| Hi-index | 0.00 |
A synergetic classification algorithm for lymphocyte discrimination is developed. By improving the ISODATA algorithm, it is possible to cluster the cell types in the training set with respect to the cluster center samples which plays a role as the cell prototype set types. Based on synergetic theory, a two-step synergetic competition mechanism was established: (1) the order parameter competition was introduced between the test cell sample and the prototypes of each prototype set, by setting the prototype of the maximum order parameter as the best prototype of each set; (2) similar matching competition was performed between the test cell sample and the best prototypes. After that the classification is achieved by selecting the most similar result as the final result. This classification method was applied on four cell groups, i.e. mantle cells, follicular centrocytes, follicular centroblastic cells and centroblastic lymphoma cells. The results show that the averaged lymphocyte classification accuracy is up to 94.1%