Tratamiento de datos para análisis de procesos de extrusión de gomas.

Laburpena

The increasing automatization of the fabrication process in the industry has generated big quantities of data. This data can give us information and knowledge about a particular process; however, a little advantage has been taken of this situation. The Data Mining consists, exactly, in taking out the knowledge that exist in the big data bases generated for some process by the Artificial Inteligence techniques, it is in this field that this work has been developed. In the process of rubber extrusion techniques of processing data and dimentional reduction, like for example SAMMON, PCA and clusters has been applied. This is with the final idea of discovering clusters, patterns of conduct, outliers, etc. Those techniques are very usefull for the analysis of multivariety-process because they allow you to see the data in 2D, that led to maximize the preservation of the original structure of data. The way you take care of the data it’s crucial in the process of Data Mining and it’s necessary a good knowledge of the process. This is because in the part where you reduce the quantity of information by putting away what is repetitive and has a little value for a farther analysis.