ABSTRACT
Visual data mining techniques have proven to be of high value in exploratory data analysis and they also have a high potential for exploring large databases [31]. This is particularly important in a context where an expert user could make use of domain knowledge to either confirm or correct a dubious classification result. An example of this interactive process is presented in [83], where the graphical interactive approaches to machine learning make the learning process explicit by visualizing the data and letting the user ‘draw’ the decision boundaries. In this work, parameters and model selection are no longer required because the user controls every step of the inductive process.
