DocumentCode
352929
Title
Characteristics of small scale nonmonotonic neuron networks having large potentiality for learning
Author
Kinjo, Mitsunaga ; Sato, Shigeo ; Nakajima, Koji
Author_Institution
Res. Inst. of Electr. Commun., Tohoku Univ., Sendai, Japan
Volume
4
fYear
2000
fDate
2000
Firstpage
171
Abstract
We report a study on learning ability of a deterministic Boltzmann machine (DBM) with neurons which have a nonmonotonic activation function. We use an end-cut-off-type function with a threshold parameter `θ´ as the nonmonotonic function. Numerical simulations of nonlinear problems, such as the 2-parity problem and the 4-parity problem, show that the DBM network with nonmonotonic neurons has higher learning ability compared to the network with monotonic neurons
Keywords
Boltzmann machines; learning (artificial intelligence); probability; 2-parity problem; 4-parity problem; deterministic Boltzmann machine; learning ability; nonmonotonic neuron networks; probability; Biomembranes; Convergence; Information processing; Intelligent networks; Intelligent systems; Laboratories; Learning systems; Machine learning; Neurons; Numerical simulation;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on
Conference_Location
Como
ISSN
1098-7576
Print_ISBN
0-7695-0619-4
Type
conf
DOI
10.1109/IJCNN.2000.860768
Filename
860768
Link To Document