The categorical criteria editor is used for the following criteria types when they have no more than eight groupings:
- Symbolic modeled analysis criteria
- Numeric and symbolic static analysis criteria
The following information is available:
- Performance - The predictive power of this criterion.
- The percentage of the population within the group, represented by the length of each bar. Percentage values are displayed as labels on the bottom horizontal axis.
The propensity of individual groups, represented by the red dots. Propensity is the ratio of the number of positive responses to the total number of responses. The propensity values are mapped to the top horizontal axis. This information is only available for modeled analysis criteria.
Range - Symbols in this group. To see this information, move your mouse over a bar or a propensity dot. The information is displayed as a tool tip.
- People (%) - Percentage of population (based on Sample data) in this group. To see this information, move your mouse over a column. The information is displayed as a tool tip.
The predictive power of this group. To see this information, move your mouse over a propensity dot.
Click on a bar to select or deselect groupings. You can select multiple groupings, including Missing. For modeled analysis criteria, the system selects the bar with the highest propensity by default.
Pega Customer Decision Hub uses the symbols associated with each group to compose suitable queries for this criterion. When multiple symbols exist in a group, the system will query for all present symbols. For the Missing group, the system translates the query to search for where the criterion value is null.
Previous topic Histograms Next topic Long Categoricals