Developing a New Suit Sizing System Using Neural Network

Manufacturing garments has the highest added value in the textile industry. Presently, most sizing systems are outdated, so each country needs to develop a new sizing system for its people. The main goal of this work was to present the new sizing chart for Iranian male suites using an artificial neural network. 10 effective sizing variables for producing suits are determined, and then all different body sizes are clustered using the Kohonen neural network. Aggregate loss is used as a tool to measure fitness of the newly obtained sizing chart. The results have a good coverage on Iranian body types, and could be used as a reference for apparel companies in Iran and other companies which produce suit for Iranian males.