Optimal solution identification system and method
View Patent ↗A method, computer program product, and computer system for receiving, at a computing device, data from at least one user, wherein at least a portion of the data is associated with the plurality of attribute variants. A variant relationship between one or more pairs of attribute variants of the plurality of attribute variants is determined based upon, at least in part, at least the portion of the data. The distance between the one or more pairs of objects is adjusted based upon, at least in part, the variant relationship between the one or more pairs of attribute variants of the plurality of attribute variants.
1. A computer-implemented method for performing analysis on a set of objects having a plurality of attributes and a related distance metric between pairs of objects, the method comprising:
calculating, at a computing device, utility data associated with attributes of a first type and a second type;
determining, at the computing device, a relationship between the first type and the second type of attributes by calculating a first difference metric between the utility values for the first type of attribute, and calculating a second difference metric between the utility values for the second type of attribute; and
determining, at the computing device, a relative attribute importance based on which of the first type of attribute or the second type of attribute affects the pairs of objects to a greater degree, wherein the relative attribute importance is based on a first within-attribute dispersion of a variant batting average of the first type of attribute and a second within-attribute dispersion of a variant batting average of the second type of attribute;
determining, at the computing device, a correlation between the first type of attribute and the second type of attribute;
determining, at the computing device, a variant similarity by transforming the correlation between the first type of attribute and the second type of attribute, wherein the correlation between the first type of attribute and the second type of attribute is transformed into the variant similarity by taking one over one plus e to the negative power of the correlation times a scale parameter plus a location parameter; and
improving a search efficiency computation by adjusting, at the computing device, the related distance metric between the pairs of objects in the set based upon the relative attribute importance and the variant similarity.
2. The computer-implemented method of claim 1 wherein the utility data includes at least one of choice data or rating data.
3. The computer-implemented method of claim 1 wherein the plurality of attributes is based upon one or more categorical variables.
4. The computer-implemented method of claim 1 wherein the plurality of attributes is based upon one or more continuous variables.
5. The computer-implemented method of claim 1 wherein the correlation between the first type and the second type of attributes is determined across a first user and a second user.
6. The computer-implemented method of claim 1 wherein the adjusting of the related distance metric between the one or more pairs of objects includes reducing a distance metric between positively correlated first type or second type of attributes.
7. The computer-implemented method of claim 1 further including clustering the set of objects based upon the adjusted related distance metric between the pairs of objects in the set of objects.
8. A computing system for performing analysis on a set of objects having a plurality of attributes and a related distance metric between pairs of objects, the computing system including a processor and a memory configured to:
calculate utility data associated with attributes of a first type and a second type;
determine a relationship between the first type and the second type of attributes by calculating a first difference metric between the utility values for the first type of attribute and calculating a second difference metric between the utility values for the second type of attribute; and
determine a relative attribute importance based on which of the first type of attribute or the second type of attribute affects the pairs of objects to a greater degree, wherein the relative attribute importance is based on a first within-attribute dispersion of a variant batting average of the first type of attribute and a second within-attribute dispersion of a variant batting average of the second type of attribute;
determine a correlation between the first type of attribute and the second type of attribute;
determine a variant similarity by transforming the correlation between the first type of attribute or the second type of attribute, wherein the correlation between the first type of attribute and the second type of attribute is transformed into the variant similarity by taking one over one plus e to the negative power of the correlation times a scale parameter plus a location parameter; and
improve a search efficiency computation by adjusting the related distance metric between the pairs of objects in the set based upon the relative attribute importance and the variant similarity.
9. The computing system of claim 8 wherein the utility data includes at least one of choice data or rating data.
10. The computing system of claim 8 wherein the plurality of attributes is based upon one or more categorical variables.
11. The computing system of claim 8 wherein the plurality of attributes is based upon one or more continuous variables.
12. The computing system of claim 8 wherein the correlation between the first type and the second type of attributes is determined across a first user and a second user.
13. The computing system of claim 8 wherein the adjusting of the related distance metric between the one or more pairs of objects includes reducing a distance metric between positively correlated attributes of the first type or the second type.
14. The computing system of claim 8 further including clustering the set of objects based upon the adjusted related distance metric between the pairs of objects in the set of objects.
15. A computer readable storage medium comprising instructions which, when executed by a processor, cause the processor to, at least:
calculate utility data associated with attributes of a first type and a second type;
determine a relationship between the first type and the second type of attributes by calculating a first difference metric between the utility values for the first type of attribute, and calculating a second difference metric between the utility values for the second type of attribute; and
determine a relative attribute importance based on which of the first type of attribute or the second type of attribute affects pairs of objects to a greater degree, wherein the relative attribute importance is based on a first within-attribute dispersion of a variant batting average of the first type of attribute and a second within-attribute dispersion of a variant batting average of the second type of attribute;
determine a correlation between the first type of attribute and the second type of attribute;
determine a variant similarity by transforming the correlation between the first type of attribute or the second type of attribute, wherein the correlation between the first type of attribute and the second type of attribute is transformed into the variant similarity by taking one over one plus e to the negative power of the correlation times a scale parameter plus a location parameter; and
improve a search efficiency computation by adjusting the related distance metric between the pairs of objects in the set based upon the relative attribute importance and the variant similarity.
16. The computer readable storage medium of claim 15 wherein the utility data includes at least one of choice data or rating data.
17. The computer readable storage medium of claim 15 wherein the plurality of attributes is based upon one or more categorical variables.
18. The computer readable storage medium of claim 15 wherein the plurality of attributes is based upon one or more continuous variables.
19. The computer readable storage medium of claim 15 wherein the correlation between the first type and the second type of attributes is determined across a first user and a second user.
20. The computer readable storage medium of claim 15 wherein the adjusting of the related distance metric between the one or more pairs of objects includes reducing a distance metric between positively correlated attributes of the first or second type.
21. The computer readable storage medium of claim 15 further including clustering the set of objects based upon the adjusted related distance metric between the pairs of objects in the set of objects.