Explanatory Tools for Machine Learning in the Symbolic Data Analysis Framework
Edwin Diday · 2020
This chapter presents mainly to give explanatory tools for the understanding of standard, complex and big data. Symbolic Data Analysis (SDA) is an extension of standard data analysis and data mining to symbolic data. The chapter defines a theoretical framework for SDA in the case of bar chart symbolic variables. Tools are given for ranking individuals, classes and their symbolic descriptive variables from the more toward the less characteristic. Classes obtained by clustering or a priori given in unsupervised or supervised learning machine are considered as new units to be described in their main facets and to be studied by taking care of their internal variability. The chapter provides various kinds of explanatory power of clustering methods based on the symbolic data table that they induce by aggregation. It also presents two clustering tools: dynamic clustering method and by mixture decomposition with the estimation-maximization method.