Lihua Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Lihua Yang has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Chao Huang
Shenzhen Univ, Coll Math & Stat, Shenzhen 518060, Peoples R China
11 shared publications Zhihua Yang
7 shared publicationsFeng Zhou
6 shared publicationsLijun Yang
5 shared publications- Zhihua Yang
Guangdong Univ Finance & Econ, Sch Informat Sci, Guangzhou 510320, Peoples R China
5 shared publications - Feng Zhou
Guangdong University of Finance and Economics
4 shared publications
Latest Publications
- 19 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 10, 332-346
Piecewise-Constant Representation and Sampling of Bandlimited Signals on Graphs
Signal representations on graphs are at the heart of most graph signal processing techniques, allowing for targeted signal models for tasks such as denoising, compression, sampling, reconstruction and detection. This paper studies the piecewise-constant representation of bandlimited graph signals, thereby establishing the relationship between the bandlimited graph signal and the piecewise-constant graph signal. For this purpose, we first introduce the concept of $\epsilon$ -level piecewise-constant representation for a general signal space. Then, using a distance matrix, a single-layer piecewise-constant representation algorithm is proposed to find an $\epsilon$ -level piecewise-constant representation for bandlimited graph signals. On this basis, we further propose a multi-layer piecewise-constant representation algorithm, which can find a node partition with as few pieces as possible to represent bandlimited graph signals piecewise within a preset error bound. Finally, as an application, we apply the node partition obtained by the multi-layer algorithm to establish a sampling theory for bandlimited signals, which does not need to compute the eigendecomposition of a variation operator in both sampling and signal reconstruction. Numerical experiments show that the proposed algorithms have good performance.
- 19 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 10, 332-346
Piecewise-Constant Representation and Sampling of Bandlimited Signals on Graphs
Signal representations on graphs are at the heart of most graph signal processing techniques, allowing for targeted signal models for tasks such as denoising, compression, sampling, reconstruction and detection. This paper studies the piecewise-constant representation of bandlimited graph signals, thereby establishing the relationship between the bandlimited graph signal and the piecewise-constant graph signal. For this purpose, we first introduce the concept of $\epsilon$ -level piecewise-constant representation for a general signal space. Then, using a distance matrix, a single-layer piecewise-constant representation algorithm is proposed to find an $\epsilon$ -level piecewise-constant representation for bandlimited graph signals. On this basis, we further propose a multi-layer piecewise-constant representation algorithm, which can find a node partition with as few pieces as possible to represent bandlimited graph signals piecewise within a preset error bound. Finally, as an application, we apply the node partition obtained by the multi-layer algorithm to establish a sampling theory for bandlimited signals, which does not need to compute the eigendecomposition of a variation operator in both sampling and signal reconstruction. Numerical experiments show that the proposed algorithms have good performance.
- 9 March 2024
- Published by World Scientific Pub Co Pte Ltd in Analysis and Applications
- Vol. 22 (05) , 937-964
Smoothing algorithms for nonsmooth optimization over the Stiefel manifold with applications to the graph Fourier basis problem
In this paper, we consider a class of nonsmooth and nonconvex optimization problems over the Stiefel manifold where the objective function is the summation of a nonconvex smooth function and a nonsmooth Lipschitz continuous convex function composed with a linear mapping. Besides, we are interested in its application to the graph Fourier basis problem. We propose three numerical algorithms for solving this problem, by combining smoothing methods and some existing algorithms for smooth optimization over the Stiefel manifold. In particular, we approximate the aforementioned nonsmooth convex function by its Moreau envelope in our smoothing methods, and prove that the Moreau envelope has many favorable properties. Thanks to this and the scheme for updating the smoothing parameter, we show that any accumulation point of the solution sequence generated by the proposed algorithms is a stationary point of the original optimization problem. Numerical experiments on building graph Fourier basis are conducted to demonstrate the efficiency of the proposed algorithms.
- 9 March 2024
- Published by World Scientific Pub Co Pte Ltd in Analysis and Applications
- Vol. 22 (05) , 937-964
Smoothing algorithms for nonsmooth optimization over the Stiefel manifold with applications to the graph Fourier basis problem
In this paper, we consider a class of nonsmooth and nonconvex optimization problems over the Stiefel manifold where the objective function is the summation of a nonconvex smooth function and a nonsmooth Lipschitz continuous convex function composed with a linear mapping. Besides, we are interested in its application to the graph Fourier basis problem. We propose three numerical algorithms for solving this problem, by combining smoothing methods and some existing algorithms for smooth optimization over the Stiefel manifold. In particular, we approximate the aforementioned nonsmooth convex function by its Moreau envelope in our smoothing methods, and prove that the Moreau envelope has many favorable properties. Thanks to this and the scheme for updating the smoothing parameter, we show that any accumulation point of the solution sequence generated by the proposed algorithms is a stationary point of the original optimization problem. Numerical experiments on building graph Fourier basis are conducted to demonstrate the efficiency of the proposed algorithms.
- Open Access
An Equivalent Graph Reconstruction Model and its Application in Recommendation Prediction
Recommendation algorithm plays an important role in recommendation system (RS), which predicts users' interests and preferences for some given items based on their known information. Recently, a recommendation algorithm based on the graph Laplacian regularization was proposed, which treats the prediction problem of the recommendation system as a reconstruction issue of small samples of the graph signal under the same graph model. Such a technique takes into account both known and unknown labeled samples information, thereby obtaining good prediction accuracy. However, when the data size is large, solving the reconstruction model is computationally expensive even with an approximate strategy. In this paper, we propose an equivalent reconstruction model that can be solved exactly with extremely low computational cost. Finally, a final prediction algorithm is proposed. We find in the experiments that the proposed method significantly reduces the computational cost while maintaining a good prediction accuracy.
Open AccessAn Equivalent Graph Reconstruction Model and its Application in Recommendation Prediction
Recommendation algorithm plays an important role in recommendation system (RS), which predicts users' interests and preferences for some given items based on their known information. Recently, a recommendation algorithm based on the graph Laplacian regularization was proposed, which treats the prediction problem of the recommendation system as a reconstruction issue of small samples of the graph signal under the same graph model. Such a technique takes into account both known and unknown labeled samples information, thereby obtaining good prediction accuracy. However, when the data size is large, solving the reconstruction model is computationally expensive even with an approximate strategy. In this paper, we propose an equivalent reconstruction model that can be solved exactly with extremely low computational cost. Finally, a final prediction algorithm is proposed. We find in the experiments that the proposed method significantly reduces the computational cost while maintaining a good prediction accuracy.
- 6 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 9, 207-216
Spline-Like Wavelet Filterbanks With Perfect Reconstruction on Arbitrary Graphs
In this work, we propose a class of spline-like wavelet filterbanks for graph signals. These filterbanks possess the properties of critical sampling and perfect reconstruction. The analysis filters are localized in the graph domain because they are polynomials in the normalized adjacency matrix of the graph. We generalize the spline-like filters in the literature so that the lowpass filter and the highpass filter can respectively remove the $s$ highest frequency components and the $r$ lowest frequency components of the signal, where $r$ and $s$ are hyperparameters specified by the users. Optimization models are formulated for the analysis filters to approximate the desired responses. Experimental results demonstrate the good locality and denoising ability of the proposed filterbanks.
- 6 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 9, 207-216
Spline-Like Wavelet Filterbanks With Perfect Reconstruction on Arbitrary Graphs
In this work, we propose a class of spline-like wavelet filterbanks for graph signals. These filterbanks possess the properties of critical sampling and perfect reconstruction. The analysis filters are localized in the graph domain because they are polynomials in the normalized adjacency matrix of the graph. We generalize the spline-like filters in the literature so that the lowpass filter and the highpass filter can respectively remove the $s$ highest frequency components and the $r$ lowest frequency components of the signal, where $r$ and $s$ are hyperparameters specified by the users. Optimization models are formulated for the analysis filters to approximate the desired responses. Experimental results demonstrate the good locality and denoising ability of the proposed filterbanks.
- Open Access
Smoothing algorithms for nonsmooth and nonconvex minimization over the stiefel manifold
We consider a class of nonsmooth and nonconvex optimization problems over the Stiefel manifold where the objective function is the summation of a nonconvex smooth function and a nonsmooth Lipschitz continuous convex function composed with an linear mapping. We propose three numerical algorithms for solving this problem, by combining smoothing methods and some existing algorithms for smooth optimization over the Stiefel manifold. In particular, we approximate the aforementioned nonsmooth convex function by its Moreau envelope in our smoothing methods, and prove that the Moreau envelope has many favorable properties. Thanks to this and the scheme for updating the smoothing parameter, we show that any accumulation point of the solution sequence generated by the proposed algorithms is a stationary point of the original optimization problem. Numerical experiments on building graph Fourier basis are conducted to demonstrate the efficiency of the proposed algorithms.
Open AccessSmoothing algorithms for nonsmooth and nonconvex minimization over the stiefel manifold
We consider a class of nonsmooth and nonconvex optimization problems over the Stiefel manifold where the objective function is the summation of a nonconvex smooth function and a nonsmooth Lipschitz continuous convex function composed with an linear mapping. We propose three numerical algorithms for solving this problem, by combining smoothing methods and some existing algorithms for smooth optimization over the Stiefel manifold. In particular, we approximate the aforementioned nonsmooth convex function by its Moreau envelope in our smoothing methods, and prove that the Moreau envelope has many favorable properties. Thanks to this and the scheme for updating the smoothing parameter, we show that any accumulation point of the solution sequence generated by the proposed algorithms is a stationary point of the original optimization problem. Numerical experiments on building graph Fourier basis are conducted to demonstrate the efficiency of the proposed algorithms.
- 19 July 2022
- Published by Elsevier in Digital Signal Processing
- Vol. 129
Atomic filter: A weak form of shift operator for graph signals
- 19 July 2022
- Published by Elsevier in Digital Signal Processing
- Vol. 129
Atomic filter: A weak form of shift operator for graph signals
- 2 June 2022
- Published by Elsevier in Neural Networks
- Vol. 153, 49-63
Approximation properties of Gaussian-binary restricted Boltzmann machines and Gaussian-binary deep belief networks
- 2 June 2022
- Published by Elsevier in Neural Networks
- Vol. 153, 49-63
Approximation properties of Gaussian-binary restricted Boltzmann machines and Gaussian-binary deep belief networks
- 9 April 2022
- Published by Elsevier in Expert Systems with Applications
- Vol. 201
A recommendation prediction method based on the estimation of PSD of sampled signals on graph
- 9 April 2022
- Published by Elsevier in Expert Systems with Applications
- Vol. 201
A recommendation prediction method based on the estimation of PSD of sampled signals on graph
- 1 November 2021
- Published by Elsevier in Signal Processing
- Vol. 188
An orthogonal partition selection strategy for the sampling of graph signals with successive local aggregations
- 1 November 2021
- Published by Elsevier in Signal Processing
- Vol. 188
An orthogonal partition selection strategy for the sampling of graph signals with successive local aggregations