Standard data mining techniques no longer adequately represent the complexity of the world. So, a new paradigm is necessary. Symbolic Data Analysis is a new type of data analysis that allows us to represent the complexity of reality, maintaining the internal variation and structure developed by Diday (2003). This new paradigm is based on the concept of symbolic object, which is a mathematical model of a concept. In this article the authors are going to present the fundamentals of the symbolic data analysis paradigm and the symbolic object concept. Theoretical aspects and examples allow the authors to understand the SDA paradigm as a tool for mining complex data.
Mixture decomposition of distributions by copulas in the symbolic data analysis framework
