Nt1210 Equation 5.4

For the output layer, the input values oi, and the output values oo (also denoted by y) are given by equation 5.4 and 5.5. ξk in equation 5.4 is a threshold value which is also adapted during the training procedure. Equation 5.5 sigmoid transfer function is applied. oi_k=∑_(j=1)^(n hidden)▒ who_(j,k).ho_j+ξ_k (5.4) y_k=〖oo〗_k=1/(1+exp⁡(-〖oi〗_k)) (5.5) where, Ninputs = Number of input parameters of the problems
Nhidden = Number of units of the hidden layer.
Nclasses = Number of classes of the problems wihij represents to the value of the weighted connection

The first set is the training set which is used for adopting the weights whereas the second set is the validation set which is used for selecting the network configuration with the best estimated generalization capability. The third set is the test which is used for obtaining the accuracy of the network and the final output.
The Nonlinear AutoRegressive model process with eXogenous input (NARX) is proposed for wind speed forecasting. The main aim of this experiment is to forecast wind speed with meteorological time series data as input variable using NARX model. Prior to forecasting, Enhanced-RReliefF feature selection is used for identification of important features for wind speed forecast and reduces the complexity of the model. Performance is evaluated in terms of mean square error when using the feature selection method with the NARX model. The framework of the proposed wind speed architecture is shown in figure