Please use this identifier to cite or link to this item: http://nopr.niscair.res.in/handle/123456789/28913
Title: Comparison of artificial neural network and regression models for prediction of air-jet textured yarn properties
Authors: Baldua, R K
Rengasamy, R S
Kothari, V K
Keywords: Air-jet textured;Artificial neural network;Instability;Loss in tenacity;Physical bulk;Regression model
Issue Date: Jun-2014
Publisher: NISCAIR-CSIR, India
Abstract: Artificial neural network (ANN) model has been designed to predict the air-jet textured yarn properties and the performance of ANN model is compared with response surface model based on multiple non-linear regression analysis. Linear density per filament, overfeed, air-pressure and texturing-speed have been selected as input variables as they have significant influence on yarn properties. Artificial neural network is able to forecast the air-jet textured yarn properties based on selected input parameters with a lower level of errors than the regression models. The validation data set, which is used to validate both the model, shows lower level of mean error per cent in case of ANN than in case of regression model.
Page(s): 157-162
URI: http://hdl.handle.net/123456789/28913
ISSN: 0975-1025 (Online); 0971-0426 (Print)
Appears in Collections:IJFTR Vol.39(2) [June 2014]

Files in This Item:
File Description SizeFormat 
IJFTR 39(2) 157-162.pdf139.92 kBAdobe PDFView/Open


Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.