Top-$k$ Regularization for Supervised Feature Selection

Feature selection identifies subsets of informative features and reduces dimensions in the original feature space, helping provide insights into data generation or a variety of domain problems. Existing methods mainly depend on feature scoring functions or sparse regularizations; nonetheless, they have limited ability to reconcile the representativeness and inter-correlations of features. In this paper, we introduce a novel, simple yet effective regularization approach, named top-$k$ regularization, to supervised feature selection in regression and classification tasks. Structurally, the top-$k$ regularization induces a sub-architecture on the architecture of a learning model to boost its ability to select the most informative features and model complex nonlinear relationships simultaneously. Theoretically, we derive and mathematically prove a uniform approximation error bound for using this approach to approximate high-dimensional sparse functions. Extensive experiments on a wide variety of benchmarking datasets show that the top-$k$ regularization is effective and stable for supervised feature selection.

Hyperspectral Image Denoising with Log-Based Robust PCA

It is a challenging task to remove heavy and mixed types of noise from Hyperspectral images (HSIs). In this paper, we propose a novel nonconvex approach to RPCA for HSI denoising, which adopts the log-determinant rank approximation and a novel $\ell_{2,\log}$ norm, to restrict the low-rank or column-wise sparse properties for the component matrices, respectively.For the $\ell_{2,\log}$-regularized shrinkage problem, we develop an efficient, closed-form solution, which is named $\ell_{2,\log}$-shrinkage operator, which can be generally used in other problems. Extensive experiments on both simulated and real HSIs demonstrate the effectiveness of the proposed method in denoising HSIs.

Deepened Graph Auto-Encoders Help Stabilize and Enhance Link Prediction

Graph neural networks have been used for a variety of learning tasks, such as link prediction, node classification, and node clustering. Among them, link prediction is a relatively under-studied graph learning task, with current state-of-the-art models based on one- or two-layer of shallow graph auto-encoder (GAE) architectures. In this paper, we focus on addressing a limitation of current methods for link prediction, which can only use shallow GAEs and variational GAEs, and creating effective methods to deepen (variational) GAE architectures to achieve stable and competitive performance. Our proposed methods innovatively incorporate standard auto-encoders (AEs) into the architectures of GAEs, where standard AEs are leveraged to learn essential, low-dimensional representations via seamlessly integrating the adjacency information and node features, while GAEs further build multi-scaled low-dimensional representations via residual connections to learn a compact overall embedding for link prediction. Empirically, extensive experiments on various benchmarking datasets verify the effectiveness of our methods and demonstrate the competitive performance of our deepened graph models for link prediction. Theoretically, we prove that our deep extensions inclusively express multiple polynomial filters with different orders.

Path Loss Modeling and Measurements for Reconfigurable Intelligent Surfaces in the Millimeter-Wave Frequency Band

Reconfigurable intelligent surfaces (RISs) provide an interface between the electromagnetic world of the wireless propagation environment and the digital world of information science. Simple yet sufficiently accurate path loss models for RISs are an important basis for theoretical analysis and optimization of RIS-assisted wireless communication systems. In this paper, we refine our previously proposed free-space path loss model for RISs to make it simpler, more applicable, and easier to use. In the proposed path loss model, the impact of the radiation patterns of the antennas and unit cells of the RIS is formulated in terms of an angle-dependent loss factor. The refined model gives more accurate estimates of the path loss of RISs comprised of unit cells with a deep sub-wavelength size. The free-space path loss model of the sub-channel provided by a single unit cell is also explicitly provided. In addition, two fabricated RISs, which are designed to operate in the millimeter-wave (mmWave) band, are utilized to carry out a measurement campaign in order to characterize and validate the proposed path loss model for RIS-assisted wireless communications. The measurement results corroborate the proposed analytical model. The proposed refined path loss model for RISs reveals that the reflecting capability of a single unit cell is proportional to its physical aperture and to an angle-dependent factor. In particular, the far-field beamforming gain provided by an RIS is mainly determined by the total area of the surface and by the angles of incidence and reflection.

Adaptive Weighted Discriminator for Training Generative Adversarial Networks

Generative adversarial network (GAN) has become one of the most important neural network models for classical unsupervised machine learning. A variety of discriminator loss functions have been developed to train GAN's discriminators and they all have a common structure: a sum of real and fake losses that only depends on the actual and generated data respectively. One challenge associated with an equally weighted sum of two losses is that the training may benefit one loss but harm the other, which we show causes instability and mode collapse. In this paper, we introduce a new family of discriminator loss functions that adopts a weighted sum of real and fake parts, which we call adaptive weighted loss functions or aw-loss functions. Using the gradients of the real and fake parts of the loss, we can adaptively choose weights to train a discriminator in the direction that benefits the GAN's stability. Our method can be potentially applied to any discriminator model with a loss that is a sum of the real and fake parts. Experiments validated the effectiveness of our loss functions on an unconditional image generation task, improving the baseline results by a significant margin on CIFAR-10, STL-10, and CIFAR-100 datasets in Inception Scores and FID.

Kernel Two-Dimensional Ridge Regression for Subspace Clustering

Subspace clustering methods have been widely studied recently. When the inputs are 2-dimensional (2D) data, existing subspace clustering methods usually convert them into vectors, which severely damages inherent structures and relationships from original data. In this paper, we propose a novel subspace clustering method for 2D data. It directly uses 2D data as inputs such that the learning of representations benefits from inherent structures and relationships of the data. It simultaneously seeks image projection and representation coefficients such that they mutually enhance each other and lead to powerful data representations. An efficient algorithm is developed to solve the proposed objective function with provable decreasing and convergence property. Extensive experimental results verify the effectiveness of the new method.

Fractal Autoencoders for Feature Selection

Feature selection reduces the dimensionality of data by identifying a subset of the most informative features. In this paper, we propose an innovative framework for unsupervised feature selection, called fractal autoencoders (FAE). It trains a neural network (NN) to pinpoint informative features for global exploring of representability and for local excavating of diversity. Architecturally, FAE extends autoencoders by adding a one-to-one scoring layer and a small sub-NN for feature selection in an unsupervised fashion. With such a concise architecture, FAE achieves state-of-the-art performances; extensive experimental results on fourteen datasets, including very high-dimensional data, have demonstrated the superiority of FAE over existing contemporary methods for unsupervised feature selection. In particular, FAE exhibits substantial advantages on gene expression data exploration, reducing measurement cost by about 15% over the widely used L1000 landmark genes. Further, we show that the FAE framework is easily extensible with an application.

A Uniformly Stable Algorithm For Unsupervised Feature Selection

High-dimensional data presents challenges for data management. Feature selection, as an important dimensionality reduction technique, reduces the dimensionality of data by identifying an essential subset of input features, and it can provide interpretable, effective, and efficient insights for analysis and decision-making processes. Algorithmic stability is a key characteristic of an algorithm in its sensitivity to perturbations of input samples. In this paper, first we propose an innovative unsupervised feature selection algorithm. The architecture of our algorithm consists of a feature scorer and a feature selector. The scorer trains a neural network (NN) to score all the features globally, and the selector is in a dependence sub-NN which locally evaluates the representation abilities to select features. Further, we present algorithmic stability analysis and show our algorithm has a performance guarantee by providing a generalization error bound. Empirically, extensive experimental results on ten real-world datasets corroborate the superior generalization performance of our algorithm over contemporary algorithms. Notably, the features selected by our algorithm have comparable performance to the original features; therefore, our algorithm significantly facilitates data management.

Structured Graph Learning for Clustering and Semi-supervised Classification

Graphs have become increasingly popular in modeling structures and interactions in a wide variety of problems during the last decade. Graph-based clustering and semi-supervised classification techniques have shown impressive performance. This paper proposes a graph learning framework to preserve both the local and global structure of data. Specifically, our method uses the self-expressiveness of samples to capture the global structure and adaptive neighbor approach to respect the local structure. Furthermore, most existing graph-based methods conduct clustering and semi-supervised classification on the graph learned from the original data matrix, which doesn't have explicit cluster structure, thus they might not achieve the optimal performance. By considering rank constraint, the achieved graph will have exactly $c$ connected components if there are $c$ clusters or classes. As a byproduct of this, graph learning and label inference are jointly and iteratively implemented in a principled way. Theoretically, we show that our model is equivalent to a combination of kernel k-means and k-means methods under certain condition. Extensive experiments on clustering and semi-supervised classification demonstrate that the proposed method outperforms other state-of-the-art methods.

Two-Dimensional Semi-Nonnegative Matrix Factorization for Clustering

In this paper, we propose a new Semi-Nonnegative Matrix Factorization method for 2-dimensional (2D) data, named TS-NMF. It overcomes the drawback of existing methods that seriously damage the spatial information of the data by converting 2D data to vectors in a preprocessing step. In particular, projection matrices are sought under the guidance of building new data representations, such that the spatial information is retained and projections are enhanced by the goal of clustering, which helps construct optimal projection directions. Moreover, to exploit nonlinear structures of the data, manifold is constructed in the projected subspace, which is adaptively updated according to the projections and less afflicted with noise and outliers of the data and thus more representative in the projected space. Hence, seeking projections, building new data representations, and learning manifold are seamlessly integrated in a single model, which mutually enhance other and lead to a powerful data representation. Comprehensive experimental results verify the effectiveness of TS-NMF in comparison with several state-of-the-art algorithms, which suggests high potential of the proposed method for real world applications.