Estimation of steel guitar strings corrosion by artificial neural network

The purpose of this work was to estimate the corrosion of steel guitar strings in the artificial sweat solution by applying artificial neural network (ANN). Measurements were conducted during 1, 2, 3 and 4 weeks. The results were used for training the most appropriate model of ANN. For evaluating the method efficiency, the comparison between measured data has been compared to values estimated by ANN. The high correlation coefficient and low mean absolute error between measured and estimated output values were observed. Therefore it can be concluded that ANN are promising tool for estimation of corrosion processes.