Back to results
US 7,127,466 B2Granted
Method for selecting node variables in a binary decision tree structure
AssigneeSAS Institute Inc.
InventorsJohn C. Brocklebank, Bruce S. Weir, Wendy Czika
Life of this patent
Estimating the term…
Abstract
A method for selecting node variables in a binary decision tree structure is provided. The binary decision tree is formed by mapping node variables to known outcome variables. The method calculates a statistical measure of the significance of each input variable in an input data set and then selects an appropriate node variable on which to base the structure of the binary decision tree using an averaged statistical measure of the input variable and any co-linear input variables of the data set.
Claims16
A computer-implemented method of selecting node variables for use in building a binary decision tree, comprising the steps of:
- (a) providing an input data set including a plurality of co-linear (providing an input data set including a plurality of co-linear input variables and an associated decision state);
- input variables and an associated decision state;
- wherein an input variable has one or more neighbor input variables;
- (b) calculating a statistical measure of the significance of each of the input variables to the associated decision state;
- wherein the statistical measure indicates how related an input variable is with respect to its associated decision state;
- (c) averaging the statistical measures for each of the input variables to form an averaged statistical measure for each input variable;
- wherein the averaging for an input variable includes using statistical measures calculated for the input variable's one or more neighbor input variables to generate an averaged statistical measure for the input variable;
- “wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier;” (variable;
- wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier;)
- (d) selecting the input variable with the largest average statistical measure;
- and
- (e) using the selected input variable as a node variable for splitting the input data set into two subsets that are used in building the binary decision tree;
- (f) for each of the two subsets in step (e), repeating steps (b) through (e);
- wherein the input variables in the input data set have an order;
- wherein the order is used to determine which of the input variables are neighbors of an input variable.
12 more elements
A computer-implemented system of selecting node variables for use in building a decision tree, wherein an input data set is provided that includes a plurality of co-linear (that includes a plurality of co-linear input variables and an associated decision state);
- input variables and an associated decision state, said system comprising:
- first software instructions to calculate a statistical measure of the significance of each of the input variables to the associated decision state;
- wherein an input variable has one or more neighbor input variables;
- wherein the statistical measure indicates how related an input variable is with respect to its associated decision state;
- second software instructions to average the statistical measures for each of the input variables to form an averaged statistical measure for each input variable;
- wherein the averaging for an input variable includes using statistical measures calculated for the input variable's one or more neighbor input variables to generate an averaged statistical measure for the input variable;
- wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier, variable;
- wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier:)
- wherein the input variable is selected that satisfies a pre-selected statistical criterion;
- and
- wherein the selected input variable is used as a node variable for splitting the input data set into two subsets that are used in building the decision tree;
- wherein the two subsets are processed by repeating for the two subsets the calculating of the statistical measures by the first software instructions and the averaging for an input variable by the second software instructions;
- wherein the input variables in the input data set have an order;
- wherein the order is used to determine which of the input variables are neighbors of an input variable.
11 more elements
A computer readable medium for storing a software program that, when executed by a processor, causes the processor to perform a method for selecting node variables for use in building a binary decision tree, the method comprising the steps of:
- (a) providing an input data set including a plurality of co-linear (providing an input data set including a plurality of co-linear input variables and an associated decision state);
- input variables and an associated decision state;
- wherein an input variable has one or more neighbor input variables;
- (b) calculating a statistical measure of the significance of each of the input variables to the associated decision state;
- wherein the statistical measure indicates how related an input variable is with respect to its associated decision state;
- (c) averaging the statistical measures for each of the input variables to form an averaged statistical measure for each input variable;
- wherein the averaging for an in put variable includes using statistical measures calculated for the input variable's one or more neighbor input variables to generate an averaged statistical measure for the input variable;
- wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier;
- (variable;
- wherein the result of averaging step smoothes the input data set a significant peak which eliminates a false identifier:)
- (d) selecting the input variable with the largest average statistical measure;
- and
- (e) using the selected input variable as a node variable for splitting the input data set into two subsets that are used in building the binary decision tree;
- (f) for each of the two subsets in step (e), repeating steps (b) through (e);
- wherein the input variables in the input data set have an order;
- wherein the order is used to determine which of the input variables are neighbors of an input variable.
13 more elements
Dependencies are inferred from the claim text.