Feature selection

Abstract

A method of feature selection applicable to both forward selection and backward elimination of features is provided. The method selects features to be used as an input for a classifier based on an estimate of the area under the ROC curve of each of the classifiers. Exemplary applications are in homecare or patient monitoring, body sensor networks, environmental monitoring, image processing and questionnaire design.

Description

technical field [0001] The present invention relates to the selection of features as input to a classifier. In particular, but not exclusively, these features represent the output of sensors in a sensor network in a home care environment, for example. Background technique [0002] Techniques for dimensionality reduction in the field of supervised machine learning have received a lot of attention. In general, there are two groups of methods: feature extraction and feature selection. In feature extraction, given features are transformed into a lower dimensional space while minimizing the loss of information. One feature extraction technique is principal component analysis (PCA), which transforms multiple correlated variables into multiple uncorrelated variables (or principal components). On the other hand, for feature selection, no new features are created. Dimensionality reduction by eliminating irrelevant and redundant features. Irrelevant (or redundant) features provid...

AI Technical Summary

Problems solved by technology

Furthermore, the high cost of collecting a large number of sample data makes it desirable to obtain efficient selection strategies to remove irrelevant and redundant features

Images