Cai Xindi, Prokhorov Danil V, Wunsch Donald C
Manuscript received September 26, 2005; revised June 25, 2006 and American Power Conversion Corporation, O'Fallon, MO 63368, USA
IEEE Trans Neural Netw. 2007 May;18(3):674-84. doi: 10.1109/TNN.2007.891685.
The winner-take-all (WTA) network is useful in database management, very large scale integration (VLSI) design, and digital processing. The synthesis procedure of WTA on single-layer fully connected architecture with sigmoid transfer function is still not fully explored. We discuss the use of simultaneous recurrent networks (SRNs) trained by Kalman filter algorithms for the task of finding the maximum among N numbers. The simulation demonstrates the effectiveness of our training approach under conditions of a shared-weight SRN architecture. A more general SRN also succeeds in solving a real classification application on car engine data.
胜者全得（WTA）网络在数据库管理、超大规模集成电路（VLSI）设计和数字处理中很有用。具有 sigmoid 传递函数的单层全连接架构上的 WTA 合成过程仍未得到充分探索。我们讨论了使用由卡尔曼滤波算法训练的同步递归网络（SRN）来完成在 N 个数中找到最大值的任务。仿真证明了我们的训练方法在共享权重 SRN 架构条件下的有效性。一个更通用的 SRN 也成功解决了汽车发动机数据的实际分类应用问题。
