Semi-Supervised Nonlinear Feature Selection on Attributed Networks

The accelerating research on attributed networks with high-dimensional node attributes in various data mining tasks highlights the significance of feature selection on networked data. In view of the availability of class labels, many feature selection methods are proposed in a semi-supervised manner as data with partial labels are more accessible to us in various scenarios. More often than not, features and labels are correlated in a nonlinear way that is more intricate than linearity. In these circumstances, vast majority of existing linear algorithms could not work well since they select features according to how well the feature can linearly explain the variance of labels. Moreover, although some methods focus on nonlinear feature selection, with the neglect of the link relations between data, they are difficult to be applied to attributed networks. In this paper, we investigate how to achieve nonlinear feature selection on attributed networks with the help of both labeled and unlabeled data. Methodologically, we propose a novel semi-supervised nonlinear framework FS-GCN based on graph convolutional networks (GCNs) to select high-quality features, which can elaborately catch the nonlinear dependency between nodal attributes and class labels. Experimental results on several real-world datasets validate the superiority of FS-GCN in terms of the quality of selected features, and its robustness in the condition of the low label rate.