Estimation of Hidden Markov Chains by a Neural Network

被引：0

作者：

Ito, Yoshifusa ^{[1
]}

Izumi, Hiroyuki ^{[2
]}

Srinivasan, Cidambi ^{[3
]}

机构：

[1] Aichi Med Univ, Sch Med, Nagakute, Aichi 4801195, Japan

[2] Aichi Gakuin Univ, Fac Policy Studies, Nisshin, Aichi 4700195, Japan

[3] Univ Kentucky, Dept Stat, Lexington, KY 40506 USA

来源：

NEURAL INFORMATION PROCESSING (ICONIP 2014), PT I | 2014年 / 8834卷

关键词：

BAYESIAN DECISION; DISCRIMINANT FUNCTIONS;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The main theme is to show that a one-hidden-layer neural network, which has learned a Bayesian discriminant function, can be used for estimating hidden Markov chains. The crucial point of the algorithm is the use of the logistic function as the activation of the output unit of the network. The network learns a single discriminant function, but converts it to the individual discriminant functions at all the steps.

引用

页码：602 / 609

页数：8

共 13 条

[1]

CHURCHILL GA, 1989, B MATH BIOL, V51, P79

[2]

Duda R. O., 1973, Pattern Classification and Scene Analysis, V3

[3] Multilayer neural networks and Bayes decision theory [J].

Funahashi, K .

NEURAL NETWORKS, 1998, 11 (02) :209-213

[4]

Ito Y, 2005, LECT NOTES COMPUT SC, V3697, P253

[5] Bayesian decision theory on three-layer neural networks [J].

Ito, Y ;

Srinivasan, C .

NEUROCOMPUTING, 2005, 63 :209-228

[6]

Ito Y, 2003, LECT NOTES COMPUT SC, V2714, P253

[7] Simultaneous Approximations of Polynomials and Derivatives and Their Applications to Neural Networks [J].

Ito, Yoshifusa .

NEURAL COMPUTATION, 2008, 20 (11) :2757-2791

[8]

Ito Y, 2008, LECT NOTES COMPUT SC, V5163, P21, DOI 10.1007/978-3-540-87536-9_3

[9]

Ito Y, 2008, LECT NOTES COMPUT SC, V4984, P238

[10]

Ito Y, 2006, LECT NOTES COMPUT SC, V4132, P350

← 1 2 →