In most cases, quality of sewn apparel products is characterized by seam performance. The durability of the seam is mainly defined by its efficiency along the seam line; therefore it is one of the most important characteristics to obtain the desired seam quality. Throughout this study, seam efficiencies of woven blended fabrics were predicted using two different methodologies, i.e. ANN and regression methods. ANN with four neurons input layer, 15 neuron hidden layer and output layer with one neuron focusing on the seam efficiency was used and compared to regression line. The input variables in both predictive modes were polyester ratios, sewing needle size, stitch density and sewing thread count. The findings of this work revealed that ANN predictive model is outperformed the multiple linear regression one with lower vales of RMSE and MBE and high R 2 values.

Throughout this study, functional, aesthetic, and dimensional stability properties of knitted blended fabrics were experimentally presented. An experimental work was introduced to examine. In order to investigate the loop length of the knitted fabrics and the polyester blending ratio on bursting, air permeability, pilling resistance and shrinkage ratio of knitted fabrics, thirty fabric samples were knitted with three loop length, five levels of polyester ratios for single jersey and rib knit structures. Analysis of variance and regression analysis were used to detect the influence of loop length and polyester content on knitted fabric properties. The findings of this study revealed that independent variables under study have a massive influence on the functional and aesthetic properties of the knitted blended fabrics.