Article
Biochemical Research Methods
Dailin Gan, Guosheng Yin, Yan Dora Zhang
Summary: Biological networks play a crucial role in understanding human diseases. This study introduces a new Gaussian Graphical Model called GR2D2 for estimating biological networks and proposes a data-augmented block Gibbs sampler algorithm. The results demonstrate that GR2D2 outperforms existing techniques in estimating precision matrices and successfully identifies cancer pathways.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Mathematics, Applied
Johannes Maly
Summary: This article addresses the problem of recovering an unknown low-rank matrix from linear measurements. By using a variational formulation and alternating minimization, the authors derive reconstruction guarantees under different measurement choices. Numerical experiments are conducted to support the validity of the theoretical considerations.
APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
(2023)
Article
Mathematical & Computational Biology
Suleyman Aydinyuz, Mustafa Asci
Summary: In this study, the coding theory for k-order Gaussian Fibonacci polynomials is rearranged to create the k-order Gaussian Fibonacci coding theory, which is based on matrices Qk, Rk, and En(k). Unlike traditional algebraic coding methods, this theory allows for the correction of matrix elements that can be infinite integers. The error detection criterion is examined for k = 2, and the method is generalized for any k with an error correction method. In the simplest case of k = 2, the method's correct capability is approximately 93.33%, surpassing all well-known correction codes. It seems that for a sufficiently large k, the probability of decoding error is almost zero.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2023)
Article
Computer Science, Information Systems
Tianjun Zhang, Hao Deng, Lin Zhang, Shengjie Zhao, Xiao Liu, Yicong Zhou
Summary: This study introduces an online camera pose correction method based on the sparse direct framework and a cascade structure optimization scheme. By minimizing the photometric errors in overlapping regions of adjacent bird's-eye-view images, the method effectively corrects the poses of cameras in a surround-view system.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Woon Huei Chai, Shen-Shyang Ho, Hiok Chai Quek
Summary: This paper studies the recovery probability of the state-of-the-art sparse recovery method YALL1 and provides a generalization of a theoretical work based on a special case of YALL1 optimization problem. The results show that not only the special case but also some other cases of YALL1 optimization problem can recover any sufficiently sparse coefficient vector under certain conditions. The trade-off parameter in YALL1 allows the recovery probability to be optimally tuned. Experimental results demonstrate the superiority of YALL1 optimization problem with primal augmented Lagrangian optimization technique in terms of speed.
INFORMATION SCIENCES
(2022)
Article
Optics
Zhenhua Li, Guili Xu, Yuehua Cheng, Zhengsheng Wang, Quan Wu, Fuju Yan
Summary: A novel variational framework is proposed in this study to correct the intensity nonuniformity in infrared images, skillfully considering the characteristics of both the latent IR image and bias field. Experimental results demonstrate the feasibility of the proposed algorithm and its ability to adapt to different nonuniform conditions compared with existing methods.
Article
Engineering, Electrical & Electronic
Sharmin Kibria, Jinsub Kim, Raviv Raich
Summary: This study focuses on joint nonlinear state estimation with multi-period measurement vectors potentially corrupted by sparse gross errors. A nonlinear sparse optimization formulation is used for joint sparse error correction and robust state estimation, exploiting the sparsity and short-term invariance of error locations. A sequential convex approximation approach is introduced to solve the nonlinear sparse optimization problem with a convergence guarantee. An identifiability-aware version of the proposed algorithm is presented to improve the accuracy of gross error localization using a necessary rank condition for identifiable gross error matrix. The efficacy of the approach is demonstrated through application to power system nonlinear state estimation in IEEE 14-bus and 118-bus networks.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Article
Computer Science, Information Systems
Qian Liu, Hao Zhang, Gaofeng Shen, Fengtong Mei
Summary: The research focuses on blind identification of LDPC codes in non-cooperative communications, proposing an optimized algorithm that enhances fault tolerance and reduces the number of iterations through proper threshold setting, optimized operations, and technique selection.
COMPUTER COMMUNICATIONS
(2021)
Article
Operations Research & Management Science
Haiyang Li, Qian Zhang, Shoujin Lin, Jigen Peng
Summary: In this paper, a method for sparse signal recovery by replacing the l(0) norm with the variation of a generalized Gaussian function Phi(alpha)(x) is proposed. The equivalence among three minimization models is demonstrated. A successful algorithm called the DCS algorithm, based on the DC algorithm and iterative soft thresholding algorithm, is presented for the regularization minimization. Experiment results show that the DCS algorithm performs well in sparse signal recovery.
JOURNAL OF GLOBAL OPTIMIZATION
(2022)
Article
Computer Science, Artificial Intelligence
Yuhao Shao, Jielin Jiang, Xiangming Hong
Summary: Noise removal is a classic problem, and mixed noise removal has always been challenging. This paper proposes an improved nonlocal means strategy to deal with mixed noise and demonstrates its effectiveness through experiments. Additionally, combining it with a convolutional neural network can further improve the removal of mixed noise.
IET IMAGE PROCESSING
(2022)
Article
Automation & Control Systems
Haojie Huang, Xin Peng, Wei Du, Steven X. Ding, Weimin Zhong
Summary: Gaussian process regression (GPR) is a popular Bayesian method for nonlinear fitting, offering predictive capability, uncertainty measurement, and interpretable structure. However, the original GPR has limited effectiveness on Big Data problems due to its heavy complexity. While several sparse GPR methods have been proposed to address this, they often sacrifice prediction accuracy. This article introduces a novel sparse GPR method with a new objective function to obtain hyperparameters in a different way, demonstrating better performance on prediction through experiments on real diesel engine and public datasets.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Mathematics, Applied
Harry Yserentant
Summary: This article discusses the estimation problem of matrix norms and Euclidean norms, and investigates the relationship between matrices and vectors under certain conditions.
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS
(2022)
Article
Statistics & Probability
Wenli Shi, Subhashis Ghosal, Ryan Martin
Summary: The paper discusses the importance of considering measurement error in estimating sparse, high-dimensional precision matrices and presents methods and results in this context. Results show that accounting for measurement error can provide empirical benefits, even when the measurement error is relatively small.
ELECTRONIC JOURNAL OF STATISTICS
(2021)
Review
Statistics & Probability
Abhirup Datta
Summary: This article reviews the connection of Nearest Neighbor Gaussian Process (NNGP) with the spatial precision matrix, as well as the diverse applications of sparse Cholesky matrices in spatial statistics, including spatial bootstrapping, simulation of Gaussian random fields on a large scale, and nonparametric mean function estimation using random forests.
WILEY INTERDISCIPLINARY REVIEWS-COMPUTATIONAL STATISTICS
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Paolo Sylos Labini, Massimo Bernaschi, Werner Nutt, Francesco Silvestri, Flavio Vella
Summary: This paper investigates ways to reorder the rows of a sparse matrix to reduce the number of nonzero blocks and cluster the nonzero elements into a few dense blocks, in order to accelerate sparse matrix multiplication.
2022 IEEE/ACM WORKSHOP ON IRREGULAR APPLICATIONS: ARCHITECTURES AND ALGORITHMS (IA3)
(2022)