One-pass training of optimal architecture auto-associative neural network for detecting ectopic beats

被引:31
作者
Clifford, G [1 ]
Tarassenko, L [1 ]
Townsend, N [1 ]
机构
[1] Univ Oxford, Dept Engn Sci, Neural Networks Res Grp, Oxford OX1 3PJ, England
关键词
Algorithms - Electrocardiography - Iterative methods - Matrix algebra - Principal component analysis;
D O I
10.1049/el:20010762
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
The authors have previously described a method for ectopic beat detection in the electrocardiogram using an auto-associative neural network. Here they present a method that utilises principal component analysis to optimse the complexity of the neural network and uses singular value decomposition to determine the initial values for the weights.
引用
收藏
页码:1126 / 1127
页数:2
相关论文
共 9 条
[1]  
Bishop C. M., 1995, NEURAL NETWORKS PATT
[2]   AUTO-ASSOCIATION BY MULTILAYER PERCEPTRONS AND SINGULAR VALUE DECOMPOSITION [J].
BOURLARD, H ;
KAMP, Y .
BIOLOGICAL CYBERNETICS, 1988, 59 (4-5) :291-294
[3]   UPDATING SINGULAR VALUE DECOMPOSITION [J].
BUNCH, JR ;
NIELSEN, CP .
NUMERISCHE MATHEMATIK, 1978, 31 (02) :111-129
[4]  
ECAR A, 1987, RECOMMENDED PRACTICE
[5]  
Golub G. H., 2013, Matrix Computations
[6]  
Golub GH., 1968, APL MAT, V13, P44, DOI 10.1007/978-3-662-39778-7_10
[7]   A REAL-TIME QRS DETECTION ALGORITHM [J].
PAN, J ;
TOMPKINS, WJ .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 1985, 32 (03) :230-236
[8]  
Ripley B. D., 1996, Pattern Recognition and Neural Networks
[9]  
TARASSENKO L, 2001, IN PRESS NEURAL PROC