(A = {({a_{ij}})_{n times n}}), (B = {({b_{ij}})_{n times n}}) represent the connect weight matrix and the delayed connection weight matrix with proper dimension.
In a neural network of size N, we assume that neuron i is connected to neuron j by a connection weight wi,j (drawn randomly from a standard normal distribution), i, j = 1,.., N, and wi,i = 0.
Two typical ANN sensitivity analysis methods, i.e., connection weight algorithm (CWA) and profile method (PM) have been applied.
Symmetry in the connection weight matrices and the boundedness of the activation functions are not required in this paper.
However, using a non-random initial connection weight algorithm and local minima avoidance and escape techniques can overcome these difficulties.
Connection weight analyses show that each preparation condition almost has the same influence on the membrane flux and selectivity.
Symmetry in the connection weight matrices and the boundedness of the activation functions are abandoned in this paper.
In this paper, structure and connection weight of a three-layer NN are optimized by genetic algorithm, and the optimized network is applied to helicopter sizing.
In the output neuron's activation function, such as the sigmoid function, an inner product of a connection weight vector with an input vector is computed.
Then the connection weight matrix of the BP network is designed for chromosomes of genetic algorithms, which is proved to optimize BP network.
Adjust the connection weight vector.
