Desarrollo de una herramienta para adquisición de datos en proyectos de datamining para un tren de laminación en caliente

Abstract

This paper solves the problem of the acquisition and data storage necessary for the development of Data mining projects. This project is developed following a methodology of data mining projects that is CRISP-DM. In this methodology a project follows some phases such as problem analysis, data analysis, data preprocess, modeling, evaluation, model development. Once analyzed the problem the objective is to identify the deviation in the width of the coil. The methodology proposes the phase of data analysis; this phase consists of activities or tasks that define the purpose of this article. This paper describes the tasks from data acquisition such as analyzing the data sources and definition of data base model. In addition, there is a description of the types of information that are used and the conversions for scale or compression. Another phases will be the exploration and the monitoring of the quality of the data; for these activities it will be necessary the development of tools that allow access to data stored to the members of the team of model development.. Both the tools and the design of the database have to allow a balance between the facility of storage and access. It is necessary to give emphasis to the importance of using methodologies for both data mining projects and software projects that will allow fixing milestones, generalize the development and facilitate the deployment minimizing the failures.