Automated determination of explanatory variables

Abstract

A determination is made of an explanatory variable with respect to an objective variable. A subset of data from data to be analyzed is created, in response to setting the objective variable to be analyzed to perform analysis. Association analysis is applied to analysis results, in response to a number of analysis runs exceeding a predetermined number. An association rule is derived for the explanatory variable from a result of the association analysis. An explanatory variable having a relevance value greater than a threshold value with the objective variable in the data to be analyzed is selected. The selected explanatory variable is scored as an input using the association rule to determine whether the explanatory variable is to be added or removed.

1 - 20 . (canceled) 
     
     
         21 . A method, comprising:
 selecting an explanatory variable having a relevance value greater than a threshold value with an objective variable in data to be analyzed; and   scoring the selected explanatory variable as an input to determine whether the explanatory variable is to be added or removed.   
     
     
         22 . The method of  claim 21 , wherein if the explanatory variable has a score exceeding a predetermined value then the explanatory variable is added. 
     
     
         23 . The method of  claim 22 , wherein if the explanatory variable is added then the added explanatory variable improves a measure of the objective variable. 
     
     
         24 . The method of  claim 22 , wherein if the explanatory variable has a score that is lower than a predetermined value then the explanatory variable is deleted. 
     
     
         25 . The method of  claim 24 , wherein:
 the objective variable is a measure of a number of events that have occurred; and   explanatory variables include at least an amount billed for the number of events, characteristics of users generating the number of events, and properties of the events.   
     
     
         26 . The method of  claim 25 , wherein a subset of the data is determined at least via decision tree analysis, wherein the number of events corresponds to a volume of telephonic communications, wherein the characteristics of the users include age, address, education level, and wherein the properties include number of lines and wireless usage. 
     
     
         27 . The method of  claim 26 , wherein the decision tree analysis is performed by performing operations in which branching of the explanatory variables which best classifies a parent node is repeated until a stopping rule based on a Gini coefficient is reached. 
     
     
         28 . A system, comprising:
 a memory; and   a processor coupled to the memory, wherein the processor performs operations, the operations comprising:   
       selecting an explanatory variable having a relevance value greater than a threshold value with an objective variable in data to be analyzed; and
 scoring the selected explanatory variable as an input to determine whether the explanatory variable is to be added or removed. 
 
     
     
         29 . The system of  claim 28 , wherein if the explanatory variable has a score exceeding a predetermined value then the explanatory variable is added. 
     
     
         30 . The system of  claim 29 , wherein if the explanatory variable is added then the added explanatory variable improves a measure of the objective variable. 
     
     
         31 . The system of  claim 29 , wherein if the explanatory variable has a score that is lower than a predetermined value then the explanatory variable is deleted. 
     
     
         32 . The system of  claim 31 , wherein:
 the objective variable is a measure of a number of events that have occurred; and   explanatory variables include at least an amount billed for the number of events, characteristics of users generating the number of events, and properties of the events.   
     
     
         33 . The system of  claim 32 , wherein a subset of the data is determined at least via decision tree analysis, wherein the number of events corresponds to telephonic communications, wherein the characteristics of the users include age, address, education level, and wherein the properties include number of lines and wireless usage. 
     
     
         34 . The system of  claim 33 , wherein the decision tree analysis is performed by performing operations in which branching of the explanatory variables which best classifies a parent node is repeated until a stopping rule based on a Gini coefficient is reached. 
     
     
         35 . A computer program product for determining an explanatory variable with respect to an objective variable, the computer program product comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code configured to perform operations, the operations comprising:
 selecting an explanatory variable having a relevance value greater than a threshold value with an objective variable in data to be analyzed; and   scoring the selected explanatory variable as an input to determine whether the explanatory variable is to be added or removed.   
     
     
         36 . The computer program product of  claim 35 , wherein if the explanatory variable has a score exceeding a predetermined value then the explanatory variable is added. 
     
     
         37 . The computer program product of  claim 36 , wherein if the explanatory variable is added then the added explanatory variable improves a measure of the objective variable. 
     
     
         38 . The computer program product of  claim 36 , wherein if the explanatory variable has a score that is lower than a predetermined value then the explanatory variable is deleted. 
     
     
         39 . The computer program product of  claim 38 , wherein:
 the objective variable is a measure of a number of events that have occurred; and   explanatory variables include at least an amount billed for the number of events, characteristics of users generating the number of events, and properties of the events.   
     
     
         40 . The computer program product of  claim 39 , wherein a subset of the data is determined at least via decision tree analysis, wherein the number of the events corresponds to a volume of telephonic communications, wherein the characteristics of the users include age, address, education level, and wherein the properties include number of lines and wireless usage.