Prediction of Fabric Hand by Neural Network

In this paper, the method that applied the neural network to the prediction method of the hand value was proposed.The cotton fabrics which made the structure of the textile and density change were woven, and the hand value was predicted by the information of the their textile structures.Also, it was examined about the validity of the prediction method. The information about the their textile structures are data about the ups and downs of warp and density, they are input data for neural network. It could be predicted comparatively precisely as a result of predicting the textile hand value which wasn't being learned in the neural network.The results that the textile hand value which made only density change was predicted could be predicted comparatively precisely, too. In addition, the estimation accuracy of prediction method by neural network is higher than that of the method by multiple regression analysis.