100%(7)7 out of 7 people found this document helpful
This preview shows page 1 - 2 out of 2 pages.
In step one, the data to be utilized for training and testing the network are gathered. Important considerations are that the particular problem is amenable to neural network arrangement and that adequate data exist and can be obtained (Omar, 2016).In step two, training data must be recognized, and a plan must be made for testing the performance of the network.The third and fourth step involves the selection of network architecture and a learningmethod. In these steps, the neural network design is developed. It includes topology selectionand determination of input and output nodes, hidden layers, and hidden nodes quantity. Theinput nodes must be based on the data sets available. According to Turban et al. (2014), afterchoosing the NN structure, a learning algorithm is identified. This is to identify connectionweight sets that are best in covering data training as well as having better predictive accuracy(Turban, 2016).