C4.1 Interactive Statistical Graphics

Examining data using graphical tools, such as histograms, bar charts, quantile plots, scatterplots and other statistical graphics, is a necessary part of any serious analysis effort. Statistical graphics are typically most useful at two stages of an analysis. Initially, plots are useful for data cleansing [cross-ref. C2.1] where unusual sets of values can be investigated, attribute distributions can be evaluated and transformations can be suggested using graphical techniques. After a data mining model has been built, similar techniques can be used to validate the results of the model, highlight unusual or atypical results and present the model in a way that facilitates understanding [cross-ref. C8.1, C9]