Predictive model variable management
View Patent ↗Models are generated using a variety of tools and features of a model generation platform. For example, in connection with a project in which a user generates a predictive model based on historical data about a system being modeled, the user is provided through a graphical user interface a structured sequence of model generation activities to be followed, the sequence including dimension reduction, model generation, model process validation, and model re-generation. A graphical user interface enables a user of a model generation tool to view and manage subgroups of variables associated with generation of a predictive model including original source variables and derived variables; variable response functions are displayed; variables are edited and transformed into new more predictive variables.
1. A machine-based method comprising:
providing a graphical user interface that enables a user of a model generation tool to view and manage subgroups of variables associated with generation of a predictive model, the variables including source variables that are associated with attributes of original data and derived variables subsequently modified from the source variables, one of the subgroups being identified as containing variables that are contributory and another of the subgroups being identified as containing variables that are non-contributory.
2. The method of claim 1 in which the derived variables include constructed variables.
3. The method of claim 1 also including enabling the user to move the variables between the subgroups using a pointer.
4. The method of claim 1 in which the user interface enables a user to view, with respect to a selected variable, its definition and its response distribution relative to an outcome variable.
5. The method of claim 1 in which only variables that are contributory within the one of the subgroups are used in the model that is generated.
6. The method of claim 1 in which each subgroup to which each variable belongs is stored persistently.
7. The method of claim 1 in which the derived variables include transformed variables.
8. The method of claim 1 in which the variables that are contributory comprise some of the source variables and some of the derived variables.
9. The method of claim 1 in which the variables that are non-contributory comprises some of the source variables and some of the derived variables.
10. A machine-based method comprising:
in connection with a project in which a user generates a predictive model based on historical data about a system being modeled, enabling the user through a graphical user interface to manage and view information about a distribution of predictor variables associated with the data, the distribution being determined based on strength of measurement of the predictor variables.
11. The method of claim 10 in which the information includes a status of a predictor variable.
12. The method of claim 10 in which the information includes a status of a class of non-predictor variables.
13. The method of claim 10 in which the user interface enables the user to point and click to cause display of information about the variable.
14. The method of claim 13 in which the information about the variable comprises at least one of: a description, a definition, a history of transformations, a response graph, and a link to the graph.
15. The method of claim 14 in which invocation of the link to the graph causes display of the probability of an event as a function of the variable.
16. The method of claim 10 also including enabling the user to cause a variable to be transformed, and including the transformed version of the variable in the information about predictor variables.
17. The method of claim 10 in which the predictor variables are grouped in the user interface, and the method also including enabling the user to move one or more than one variable from one group to another.
18. The method of claim 10 in which the graphical user interface represents filtered views of a data dictionary of an entire set of variables of the model.
19. The method of claim 10 in which the user interface enables the user to point and click to cause display of decision partition trees of a collection of the predictor variables.
20. The method of claim 10 in which the user interface enables the user to point and click to cause display of response distribution functions of a collection of the predictor values.
21. The method of claim 10 in which the strength of measurement comprises contributory and non-contributory.