Ara
Toplam kayıt 3, listelenen: 1-3
An improved version of backpropagation algorithm with effective dynamic learning rate and momentum
(2006)
An improvement of backpropagation algorithm with momentum is introduced. Local quadratic approximation of the error function is performed at every stage of the learning process and the Hessian matrix of the quadratic error ...
Statistical methods and artificial neural networks
(2006)
Artificial Neural Networks and statistical methods are applied on real data sets for forecasting, classification, and clustering problems. Hybrid models for two components are examined on different data sets; tourist arrival ...
Accelerating backpropagation using effective parameters at each step and an experimental evaluation
(Taylor & Francis LTD, 2008)
An acceleration of backpropagation algorithm with momentum (BPM) is introduced. At every stage of the learning process, local quadratic approximation of the error function is performed and the Hessian matrix of the quadratic ...