Back to results
US 2016/0110656 A1Published application
Feature selection
AssigneeYahoo! Inc.
InventorsMakoto Yamada, Hua Ouyang, Yi Chang, Avishek Saha
Life of this patent
Estimating the term…
Abstract
A novel method and/or system of feature selection is described.
Claims20
A method of feature selection comprising:
- generating from a set of signal samples a set of feature signal vectors for a set of features in which a respective signal vector in the set of feature signal vectors corresponds to a respective feature in the set of features;
- determining the following for the respective feature signal vectors in the set of feature signal vectors:
- a normalized version of Hilbert-Schmidt Independence between a respective feature signal vector of the set of feature signal vectors and an output signal array, a normalized version of Hilbert-Schmidt Independence between the respective feature signal vector of the set of feature signal vectors and the remaining feature signal vectors of the set of feature signal vectors, and a gradient per feature of the set of features using the previously determined normalized versions of Hilbert-Schmidt Independence for the set of feature signal vectors;
- and selecting a feature having the feature signal vector of the set of feature signal vectors that produces the largest gradient.
1 more element
The method of 1, wherein the features selected comprise substantially non-redundant features.
A system comprising:
- a computing device;
- the computing device to generate a set of feature signal vectors for a set of features in which a respective signal vector in the set of feature signal vectors corresponds to a respective feature in the set of features, to determine the following for the respective feature signal vectors in the set of feature signal vectors:
- a normalized version of Hilbert-Schmidt Independence between a respective feature signal vector of the set of feature signal vectors and an output signal array, a normalized version of Hilbert-Schmidt Independence between the respective feature signal vector of the set of feature signal vectors and the remaining feature signal vectors of the set of feature signal vectors, and a gradient per feature of the set of features to be determined with the previously determined normalized versions of Hilbert-Schmidt Independence for the set of feature signal vectors, and to select a feature having the feature signal vector of the set of feature signal vectors to produce the largest gradient.
A system comprising:
- means for generating a set of feature signal vectors for a set of features in which a respective signal vector in the set of feature signal vectors corresponds to a respective feature in the set of features;
- means for determining the following for the respective feature signal vectors in the set of feature signal vectors:
- a normalized version of Hilbert-Schmidt Independence between a respective feature signal vector of the set of feature signal vectors and an output signal array, a normalized version of Hilbert-Schmidt Independence between the respective feature signal vector of the set of feature signal vectors and the remaining feature signal vectors of the set of feature signal vectors, and a gradient per feature of the set of features using the previously determined normalized versions of Hilbert-Schmidt Independence for the set of feature signal vectors;
- and
- means for selecting a feature having the feature signal vector of the set of feature signal vectors that produces the largest gradient.
2 more elements
Dependencies are inferred from the claim text.