An Input Variable Selection Method for the Artificial Neural Network of Shear Stiffness of Worsted Fabrics

Abstract

Abstract:The relationship between yarn properties, fabric parameters, and shear stiffness of worsted fabrics is modeled using the soft computing technique. Because of the small number of samples, the artificial neural network model to be established must be a small-scale one. Therefore, this soft computing approach includes two stages. First, the yarn properties and fabric parameters are selected by utilizing an input variable selection method, so as to find the most relevant yarn properties and fabric parameters as the input variables to fit the small-scale artificial neural network model. The first part of this method takes the human knowledge on the shear stiffness of fabrics into account. The second part utilizes a data sensitivity criterion based on a distance method. Second, the artificial neural network model of the relationship between yarn properties, fabric parameters, and shear stiffness of fabrics is established. The results show that the artificial neural network model yields accurate prediction and a reasonably good artificial neural network model can be achieved with relatively few data points by integrating with the input variable selecting method developed in this research. The results also show that there is great potential for this research in the field of computer-assisted design in textile technology. 

Artificial Neural Network Prosperities in Textile Applications

“…Hence, the variables had clear physical meanings. Their results showed accurate prediction (up to average error of 0.209%) by the small-scale artificial neural network model and a reasonably good artificial neural network model could be achieved with relatively few data points by integrating with the input variable selecting method developed in their research (Chen et al, 2009). Needle punching is a well-known nonwoven process of converting fibrous webs into selflocking or coherent structures using barbed needles.…”

“…As a nonlinear problem, predicting the shear stiffness can be realized by an alternative modeling method, that is, by using the artificial neural network (ANN) model. Chen et al, 2009 modeled the relationship between yarn properties, fabric parameters, and shear stiffness of worsted fabrics using two stage neural network models. First, the yarn properties and fabric parameters were selected by utilizing an input variable selection method to find the most relevant yarn properties and fabric parameters as the input variables to fit the small-scale artificial neural network model.…”

Input Variable Selection for Oil Palm Plantation Productivity Prediction Model

“…Penelitian ini mengembangkan dan memperbaiki sebuah metode seleksi variabel input yang menggabungkan dua buah metode berdasarkan pada pendapat pengetahuan ahli perkebunan kelapa sawit dan juga sensitivitas data berbasis jarak [11] (selanjutnya akan disebut dengan nama metode CZCL). Model prediksi produksi tanaman kelapa sawit yang dikembangkan menggunakan jaringan syaraf tiruan feedforward dengan metode pembelajaran backpropagation.…”

“…Metode CZCL digunakan untuk melakukan seleksi variabel input pada permasalahan model prediksi kualitas tekstil [11]. Karakteristik data tekstil yang digunakan pada penelitian tersebut memiliki banyak kemiripan dengan data sampel tanah dan daun tanaman perkebunan kelapa sawit.…”