The data mining for TP film´s transmittance by using neural network

This paper presents a new computation method based on the weights of the well-trained neural network (NN) for the data mining of touch panel (TP) film´s transmittance. By using the method developed, the influence degree of each input variable to the transmittance could be obtained and then the useful influencing inputs could also be determined. In this research, the data of the transmittance of TP film with Cr and Cr₂O₃ coating are studied. The possible influencing factors including the coating target composition, the layers of coating material, the film´s thickness, the position of panel placed and the rotation speed of evaporator´s holder are collected and analyzed. The relationship between film´s transmittance and these possible influencing factors is expected to be obtained.