TY - CHAP
T1 - Multicategory Bayesian decision using a three-layer neural network
AU - Ito, Yoshifusa
AU - Srinivasan, Cidambi
PY - 2003
Y1 - 2003
N2 - We realize a multicategory Bayesian classifier by a three-layer neural network having rather a small number of hidden layer units. The state-conditional probability distributions are supposed to be multivariate normal distributions. The network has direct connections between the input and output layers. Its outputs are monotone mappings of posterior probabilities. Hence, they can be used as discriminant functions and, in addition, the posterior probabilities can be easily retrieved from the outputs.
AB - We realize a multicategory Bayesian classifier by a three-layer neural network having rather a small number of hidden layer units. The state-conditional probability distributions are supposed to be multivariate normal distributions. The network has direct connections between the input and output layers. Its outputs are monotone mappings of posterior probabilities. Hence, they can be used as discriminant functions and, in addition, the posterior probabilities can be easily retrieved from the outputs.
UR - http://www.scopus.com/inward/record.url?scp=35248827534&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=35248827534&partnerID=8YFLogxK
U2 - 10.1007/3-540-44989-2_31
DO - 10.1007/3-540-44989-2_31
M3 - Chapter
AN - SCOPUS:35248827534
SN - 3540404082
SN - 9783540404088
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 253
EP - 261
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
A2 - Kaynak, Okyay
A2 - Alpaydin, Ethem
A2 - Oja, Erkki
A2 - Xu, Lei
ER -