An almost sure central limit theorem for the overlap parameters in the Hopfield model

被引:7
作者
Gentz, B
机构
[1] Inst. für Angewandte Mathematik, Universität Zürich, CH-8057 Zürich
关键词
fluctuations; Hopfield model; overlap parameter; neural networks; Laplace's method;
D O I
10.1016/0304-4149(96)00055-5
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We consider the Hopfield model with a finite number of randomly chosen patterns above and below the critical temperature and prove an almost sure conditional central limit theorem for the vector of overlap parameters, For this purpose we analyse the almost sure asymptotic behaviour of the partition function.
引用
收藏
页码:243 / 262
页数:20
相关论文
共 10 条