GARMENT INDUSTRY: ESTIMATING OPTIMAL PRODUCTION OF NEW PRODUCTS USING FUNCTIONAL LINKED NEURAL NETWORKS TECHNIQUE

摘要

Predicting number of products to be produced by any firm is a complex phenomenon and depends on many factors with different degree of association which could be possibly mapped in a non-linear manner. Neural networks inherent non linearity because of their structure and proved to be a very good estimator for predicting non linear trends. To estimate the optimal production capacity for new products in garment industry functional linked neural network (FLANN) techniques have been used with different weight updating algorithm like least mean square, recursive least square and particle swarm optimization. Predicting accuracy has been calculated for every network and compiled to evaluate the best performing model.