Computational intelligence for classification of patterned glass manufacturing using data from a glass furnace model

The aim of this work is to verify the efficiency of Fuzzy Decision Tree (FID 3.4) and other algorithms in the classification of patterned glass in a manufacturing glass process, using Brazilian Union of Glass (UBV) database. The database “UBV” consists of the operating variables of the furnace. These Fuzzy algorithms were used as a tool to help selecting the most adequate temperature necessary to process the patterned glass, and so reduce defects in end products. For this purpose a Model of Manufactures was used to study the process of manufacturing patterned glass. The classification obtained through the Fuzzy Decision Tree was compared with the results obtained in the manufacture of patterned glass. The classifications obtained in the manufacture and in the FID 3.4 software were also compared with the classification obtained with different algorithms, such as CART (Classification and Regression Tree) and Artificial Neural Network (ANN).