Statistical modeling without manual settings
In contrast to classical data mining, the user defined manual settings in the self- acting data mining and/or with the automatic are minimal. The data fields for the analysis are determined automatically. Different data types can be combined automatically (source types). For instance, unstructured data from:
· PDF and Word files
· News Feeds
· XML Format
Missing values and outliers are recognized automatically. Redundancies are eliminated automatically.
The self-acting data miining procedure can also avoid using strongly correlating variables.
In this way the analysis models are provided automatically. The user must only specify the “Business Question”; here it is supported by the Prospero solution in a wizard.
In classical DM, 75% of the total time is spent for the modification and selection of the data. In self-acting DM this value can be reduced to 20%.
In practice, self-acting DM provides a smoother data analysis project.