TY - GEN
T1 - Multi-category bayesian decision by neural networks
AU - Ito, Yoshifusa
AU - Srinivasan, Cidambi
AU - Izumi, Hiroyuki
PY - 2008
Y1 - 2008
N2 - For neural networks, learning from dichotomous random samples is difficult. An example is learning of a Bayesian discriminant function. However, one-hidden-layer neural networks with fewer inner parameters can learn from such signals better than ordinary ones. We show that such neural networks can be used for approximating multi-category Bayesian discriminant functions when the state-conditional probability distributions are two dimensional normal distributions. Results of a simple simulation are shown as examples.
AB - For neural networks, learning from dichotomous random samples is difficult. An example is learning of a Bayesian discriminant function. However, one-hidden-layer neural networks with fewer inner parameters can learn from such signals better than ordinary ones. We show that such neural networks can be used for approximating multi-category Bayesian discriminant functions when the state-conditional probability distributions are two dimensional normal distributions. Results of a simple simulation are shown as examples.
UR - http://www.scopus.com/inward/record.url?scp=58849120120&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=58849120120&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-87536-9_3
DO - 10.1007/978-3-540-87536-9_3
M3 - Conference contribution
AN - SCOPUS:58849120120
SN - 3540875352
SN - 9783540875352
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 21
EP - 30
BT - Artificial Neural Networks - ICANN 2008 - 18th International Conference, Proceedings
T2 - 18th International Conference on Artificial Neural Networks, ICANN 2008
Y2 - 3 September 2008 through 6 September 2008
ER -