Test data set for the conveyor transport system

Published: 25-04-2020| Version 6 | DOI: 10.17632/4vcb843t76.6
Oleh Pihnastyi


Using the PiKh–model, a test data set for training the neural network is formed. Training data is presented in a separate file.. The architecture of the neural network can be arbitrary and is set by the settings file. To build the architecture of a neural network, it is necessary to determine the names of the input nodes, the names of the output nodes and set the parameters for hidden layers and the output layer. Each output layer is characterized by a name and parameters that determine the number of nodes, the type of activation function, the optimization algorithm, and the method for distributing errors between nodes. The settings file allows you to set the number of epochs during the training of the neural network, the interval between epochs when the learning results are saved (the interval of data recording on the hard disk), the error value (MSE), and the value of the task stop time for cooling the processor.