Analyzing data with high dimensionality, such as that generated by social media, requires complex algorithms. Feature selection reduces the dimensionality by selecting a subset of the most relevant features in high dimensional data. Feature selection methods can be broadly classiﬁed into supervised methods for labeled data and unsupervised methods for unlabeled data. Sparse learning has been proven to be a powerful technique for analyzing high dimensional data via embedded supervised feature selection. In recent years, increased efforts have been made to apply sparse learning to unsupervised feature selection. Due to the lack of label information, the vast majority of these algorithms generate cluster labels via clustering algorithms. Existing unsupervised feature selection methods must transform unsupervised feature selection into sparse learning based supervised feature selection with cluster labels generated by clustering algorithms. By creating a program that is able to perform the same tasks without transforming the data, unsupervised feature selection could be more accurate and efficient.
Researchers at Arizona State University have developed a novel unsupervised feature selection algorithm called Embedded Unsupervised Feature Selection (EUFS). The key innovation of this program is its ability to perform unsupervised feature selection without requiring a transformation. It directly embeds an unsupervised feature selection algorithm into a clustering algorithm via sparse learning instead of transforming it into sparse learning based supervised feature selection with cluster labels. This work extends the current state-of-the-art unsupervised feature selection and empirically demonstrates the efﬁcacy of the new algorithm. Additionally, experimental results show increased efficiency and accuracy for the feature selection process.
- Social Media
- Data Analytics
- Journalism and News Reporting
Benefits and Advantages
- Proven Results - Experimental trials demonstrate the effectiveness of EUFS.
- Increased efficiency – Able to use sparse learning on unsupervised feature selection data without transformation.
- Improved accuracy – Increases accuracy for feature selection which produces more relevant data for the user.
For more information about the inventor(s) and their research, please see
For more information about related technologies, please see