Abhijit Majumdar
Devinder Kumar Banwet
Abstract
This paper presents the comparative analysis of three quantitative methodologies namely linear regression, moving average and artificial neural network (ANN) to forecast the aggregate cotton fabric and yarn production in India. Cotton fabric and yarn production data were collected for the period from 1995-1996 to 2010-2011. The relative performance of these methodologies was compared with statistical parameters like correlation coefficient and mean absolute percentage error (MAPE). The performance of the ANN approach was found to be superior to that of statistics based approaches (linear regression and moving average).
Keywords- Artificial neural network, Cotton, Fabric, Forecasting, Textile industry, Yarn