A new Unsupervised Spectral Feature Selection Method for mixed data: A filter approach
| Autor: | J. Ariel Carrasco-Ochoa, José Fco. Martínez-Trinidad, Saúl Solorio-Fernández |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
0301 basic medicine
Computer science Feature selection 02 engineering and technology computer.software_genre Measure (mathematics) 03 medical and health sciences Search engine Artificial Intelligence 0202 electrical engineering electronic engineering information engineering business.industry Process (computing) Contrast (statistics) Pattern recognition Filter (signal processing) 030104 developmental biology Feature (computer vision) Kernel (statistics) Signal Processing 020201 artificial intelligence & image processing Computer Vision and Pattern Recognition Artificial intelligence Data mining business computer Software |
| Zdroj: | Pattern Recognition. 72:314-326 |
| ISSN: | 0031-3203 |
| Popis: | Most of the current unsupervised feature selection methods are designed to process only numerical datasets. Therefore, in practical problems, where the objects under study are described through both numerical and non-numerical features (mixed datasets), these methods cannot be directly applied. In this work, we propose a new unsupervised filter feature selection method that can be used on datasets with both numerical and non-numerical features. The proposed method is inspired by the spectral feature selection, by using together a kernel and a new spectrum based feature evaluation measure for quantifying the feature relevance. Experiments on synthetic datasets show that in the 99% of the cases where the relevant features are known our method identifies and ranks the most relevant features at the beginning of a sorted list. Additionally, we contrast our method against state-of-the-art unsupervised filter methods over real datasets, and our method in most cases significantly outperforms them. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect