Unsupervised Feature Selection Using Nonnegative Spectral Analysis

By Zechao Li, Yi Yang, J. Liu, Xiaofang Zhou, Hanqing Lu. AAAI Conference on Artificial Intelligence, 2012

541 citations recorded by the source corpus.

In this paper, a new unsupervised learning algorithm, namely Nonnegative Discriminative Feature Selection (NDFS), is proposed. To exploit the discriminative information in unsupervised scenarios, we perform spectral clustering to learn the cluster labels of the input samples, during which the feature selection is performed simultaneously. The joint learning of the cluster labels and feature selection matrix enables NDFS to select the most discriminative features. To learn more accurate cluster labels, a nonnegative constraint is explicitly imposed to the class indicators. To reduce the redundant or even noisy features, l2,1-norm minimization constraint is added into the objective function, which guarantees the feature selection matrix sparse in rows. Our algorithm exploits the discriminative information and feature correlation simultaneously to select a better feature subset. A simple yet efficient iterative algorithm is designed to optimize the proposed objective function. Experimental results on different real world datasets demonstrate the encouraging performance of our algorithm over the state-of-the-arts.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/0139eb62a87649bf7d259542b5afc6be121b094b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jiliang Tang
2. Xuelong Li
3. Huan Liu
4. Jundong Li
5. Hanqing Lu
6. Feiping Nie
7. Zechao Li
8. J. Liu

In Computer Vision and Pattern Recognition

1. Xuelong Li
2. Dacheng Tao
3. Hanqing Lu
4. Feiping Nie
5. Jinhui Tang
6. Yi Yang
7. Zechao Li
8. J. Liu

In Media Technology

1. Xuelong Li
2. Hanqing Lu
3. J. Liu
4. Zechao Li
5. Jiliang Tang
6. Huan Liu
7. Jundong Li
8. Feiping Nie

Citations in the corpus

• Robust Structured Subspace Learning for Data Representation (2015, methodology): “, Laplacian Score [17], SPEC [15], MCFS [18], UDFS [19], NDFS [46] and LLCFS [47].”
• Reconstruction-based Unsupervised Feature Selection: An Embedded Approach (2017, methodology): “Since real-world data is usually unlabeled and collecting labeled data is particular expensive requiring both time and effort, unsupervised feature selection has received increasingly attention in the past few years [He et al., 2005; Cai et al., 2010; Yang et al., 2011; Li et al., 2012; Qian and Zha”
• Pattern Recognition (2018, unknown): “To our knowledge, this is the first work to consider both identification and soft biometrics jointly for silhouette-based gait analysis.”
• Partially Supervised Graph Embedding for Positive Unlabelled Feature Selection (2016, methodology): “…compare the proposed PUFS with the other five state-of-the-art feature selection algorithms, including RFS [Nie 2010], JELSR [Chenping Hou and Wu, 2011], NDFS [Li 2012], Pairwise Ranking based Feature Selection (PRFS) [S. Sundararajan and Keerthi, 2011] and apuCFS [Borja Calvo and A.Lozano, 2009].”
• Clustering-Guided Sparse Structural Learning for Unsupervised Feature Selection (2014, methodology): “In this section, we discuss the relationships between the proposed method and several algorithms, including SPFS [39], MCFS [12], UDFS [13] and NDFS [29].”
• A review of unsupervised feature selection methods (2019, methodology): “– Multivariate: FSFS (Mitra et al. 2002), RRFS (Ferreira and Figueiredo 2012), UDFS (Yang et al. 2011b),NDFS (Li et al. 2012),UFSACO (Tabakhi et al. 2014),MGSACO (Tabakhi et al. 2015), and DSRMR (Tang et al. 2018a).”
• Weakly-Supervised Dual Clustering for Image Semantic Segmentation (2013, methodology): “To satisfy the first constraint, we impose an orthogonality constraint on Y just like [10], i.”
• Feature Selection Based on Structured Sparsity: A Comprehensive Study (2017, unknown): “[141] proposed a feature selection method”
• Unsupervised Feature Selection with Structured Graph Optimization (2016, methodology): “Li et al. (2012) jointly use FME, non-negative constraint, and -2,1 norm to perform Nonnegative Discriminative Feature Selection (NDFS).”
• Feature Selection (2016, methodology): “Nonnegative Discriminative Feature Selection (NDFS) (Li et al. 2012) performs spectral clustering and feature selection simultaneously in a joint framework to select a subset of discriminative features.”