Intelligent Engineering Systems Through Artificial Neural Networks: Proceedings of the Artificial Neural Networks in Engineering (ANNIE ... ) Conference, Volume 2ASME Press, 1992 - Intelligent control systems |
From inside the book
Results 1-3 of 82
Page 216
... matrix directly whereas the Kalman algorithm deals with the square of the data matrix , hence in the squaring operation , precision is lost . This precision can be very important in yielding an accurate weight matrix especially when the ...
... matrix directly whereas the Kalman algorithm deals with the square of the data matrix , hence in the squaring operation , precision is lost . This precision can be very important in yielding an accurate weight matrix especially when the ...
Page 306
... matrix W. Ideally , the elements in W are real numbers having infinite accuracy . However , when implementing the weight matrix in hardware , the weight accuracy is constrained by the chosen technology / circuitry . The hardware ...
... matrix W. Ideally , the elements in W are real numbers having infinite accuracy . However , when implementing the weight matrix in hardware , the weight accuracy is constrained by the chosen technology / circuitry . The hardware ...
Page 923
... matrix from input to the first hidden layer , and V , be the weight matrix from ( r - 1 ) th hidden layer to rth hidden layer . Let W be the wight matrix from last hidden layer to the output layer . Let there be n hidden layers in the ...
... matrix from input to the first hidden layer , and V , be the weight matrix from ( r - 1 ) th hidden layer to rth hidden layer . Let W be the wight matrix from last hidden layer to the output layer . Let there be n hidden layers in the ...
Contents
Design of Supervised Classifiers Using Boolean | 1 |
A Link Between | 15 |
A Neural NetworkBased Predictor for MIMO | 21 |
Copyright | |
118 other sections not shown
Other editions - View all
Common terms and phrases
accuracy activation adaptive analysis application approach Artificial Neural Networks backpropagation binary classification cluster CMAC components Computer convergence corresponding data set defined defuzzification delta rule detection distribution dynamic Engineering equation error estimate evaluation experiments fault feature feedforward feedforward neural network fuzzy fuzzy logic fuzzy set given gradient gradient descent hidden layer hidden units Hopfield network hybrid identified IEEE implementation initial input layer input pattern input vector iterations Kohonen learning algorithm learning rate linear machine mapping matrix measure method minimize multilayer multilayer perceptron neocognitron neural net neurons noise nonlinear normalized number of hidden obtained optimal output layer output node paper parameters pattern recognition perceptron performance prediction problem random represents robot rule samples selected sensor signal simulation solution space step structure supervised learning techniques texture threshold training data training set trajectory update values variables weight vector