An Algorithm, Based on Extreme Machine Learning, for Modeling Rate of Material Transfer in EDC Process

In this paper, Extreme Learning Machine method is used to model the rate of material transfer as an effective parameter in process speed and surface quality. Using neural network model of Extreme Learning Machine, the mean squared error (MSE) for the material transfer rate in the learning data is 0.000,387 and in the test data is 0.001,7. While, the mean error squared for the average reset layer thickness, calculated in the learning data, was 0.000,214 and in the test data was 0.001,7. The proposed algorithm of Extreme Learning Machine with experimental results has high accuracy in predicting a process output parameters.