Search

To address the limitations of reduced accuracy in graph diffusion methods when handling complex edge relationships, this paper proposed a curvature-based graph diffusion neural network. The method introduced Ollivier-Ricci curvature to quantify edge curvature, providing a geometric measure of graph structure. The algorithm adjusted the weights of the random transition matrix using curvature, modifying them based on geometric relationships. It then combined the processed curvature matrix with the graph diffusion matrix to update the weight coefficients for model training. Experimental results show that the improved method maintains the advantages of smoothing graph signals effectively and reducing high-frequency noise. It increased accuracy by 0. 3 to 2.0 percentage points on datasets with varying numbers of edges and nodes. The method optimized message aggregation in graph diffusion, utilizing node information and edge weights within the graph structure more effectively. This enhancement improves model performance in node classification tasks and provides a reliable method and experimental basis for future graph-based research. [ABSTRACT FROM AUTHOR]

Copyright of Space-Integrated-Ground Information Networks is the property of Beijing Xintong Media Co., Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Lubrication Engineering (0254-0150) is the property of Editorial Office of LUBRICATION ENGINEERING and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Control Theory & Applications / Kongzhi Lilun Yu Yinyong is the property of Editorial Department of Control Theory & Applications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Central China Normal University is the property of Huazhong Normal University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Blind spectrum sensing algorithms based on eigenvalues has been widely studied in multi-antenna cognitive radio networks. The design of detection algorithms based on eigenvalues relies on the random matrix theory and the characteristics of eigenvalues of sample covariance matrix. Some work has shown that high order moments of eigenvalues can provide additional identification information to improve the performance of statistical inference problems. Motivated by this, an improved spectrum sensing algorithm resorting to the pth moment of all the eigenvalues is proposed, and the analytical representations of false alarm probability and decision threshold are given using the random matrix theory. Moreover, the influence of the power change of high-order moments on the detection performance is analyzed through simulation experiments. The simulation results show the effectiveness of the proposed spectrum sensing algorithm. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Beijing University of Technology is the property of Journal of Beijing University of Technology, Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Control Theory & Applications / Kongzhi Lilun Yu Yinyong is the property of Editorial Department of Control Theory & Applications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Control Theory & Applications / Kongzhi Lilun Yu Yinyong is the property of Editorial Department of Control Theory & Applications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

To allow different organizations to jointly use PCA for dimensionality reduction and data release under the premise of protecting the privacy of local sensitive data and the data released after dimensionality reduction, this paper proposed a horizontal federated PCA differential privacy data publishing algorithm. It introduced a random seed joint negotiation scheme to generate the same noise matrix between sites with less communication cost. It proposed a local noise averaging scheme, and added the averaging noise to the local covariance matrix. On the one hand, it protected local data privacy. On the other hand, it reduced the amount of noise added and achieved the same noise level as the centralized differential privacy PCA algorithm. Compared with similar algorithms, the amount of noise added was reduced. The experiment evaluates the algorithm from the perspective of privacy and utility, and proves that the algorithm has higher utility compared with similar algorithms. [ABSTRACT FROM AUTHOR]

In high-speed cutting processes, traditional Nyquist-Shannon sampling theorem was used for data collection which confront difficult problems of storage, transmission and processing for large amount of cutting force signals. A novel method of cutting force signal acquisition was proposed to realize the compression acquisition of signals based on the ompressed sensing theory. Gauss random matrix was selected as the basic measurement matrix and was redesigned by combining the approximate orthogonal upper triangular decomposition and the minimum correlation coefficient method to improve the compression measurement performance. Then the original cutting force signals were reconstructed from the measurement values by using the efficient compressive sampling matching pursuit algorithm. The experimental results show that the improved Gauss random measurement matrix has higher reconstruction accuracy and stability, and the proposed method greatly reduce the amount of data while ensuing the reconstruction efficiency and accuracy of cutting forces. [ABSTRACT FROM AUTHOR]

To detect the target signal in composite Gaussian clutter, the clutter covariance matrix needs to be estimated. The corresponding detection performance is closely related to the estimation accuracy. Using the texture component obeying the inverse Gaussian distribution to model the composite Gaussian clutter can better fit the measured data of high-resolution clutter. In this paper, a two-step generalized likelihood ratio detector is proposed. Firstly, the clutter covariance matrix is assumed to be known to obtain the detection statistics, and then the maximum likelihood estimation of the covariance matrix is derived from the prior distribution of texture components. At the same time, based on Bayesian method, assuming that the texture component and covariance matrix are random quantities subject to a priori distribution, the maximum a posteriori estimation of covariance matrix is derived. Simulation results show that due to the introduction of prior information of texture components and clutter, the estimation accuracy of covariance matrix of knowledge-based adaptive detection technology is better than that of maximum likelihood estimation and sample estimation methods, and has better detection performance. [ABSTRACT FROM AUTHOR]

Aiming at the actual requirements of high efficiency and security in signal acquisition and transmission, a signal compression encryption method based on Chua's circuit and compressive sensing is proposed. Firstly, the SHA-256 algorithm is used to generate a message digest of the original signal; secondly, the digest is used to generate the initial value of the Chua's circuit; The chaotic observation matrix is used to compress and sample the signal by the chaotic analog signal generated from the Chua's circuit which is chosen to generate a pseudo-random sequence through 8-bit quantization acquisition and digital shift difference. Finally, the compressed signal is scrambled and encrypted with the generated pseudo-random sequence to generate a compressed and encrypted signal. The experimental results show that the performance of the chaotic observation matrix generated by this method is better than that of the traditional random observation matrix, and t can effectively improve the signal transmission efficiency and security. [ABSTRACT FROM AUTHOR]

Objectives: To solve the problem that the coefficient matrix and the observation vector have structured characteristics for many practical applications in the field of surveying and mapping, that is, the coefficient matrix and the observation vector contain fixed quantities (or even fixed columns) and random quantities, and the random quantities at different positions are linearly related.Methods: Starting from the errors-in -variables (EIV) function model, the augmented matrix composed of the coefficient matrix and the observation vector is expressed as an affine function form, and the function model is reconstructed by the variable projection method. Then, a structured total least squares (STLS) estimation algorithm is derived by the Lagrange method.Results:The example results show that the proposed algorithm is consistent with the existing weighted or structured total least squares estimation algorithms that can solve structured prob- lem in the coefficient matrix and the observation vector. Compared with the weighted total least squares (WTLS) estimation algorithms, the proposed algorithm only needs to establish a positive definite weight matrix of independent random variables and reduces the number of estimates. Compared with other STLS estimation algorithms, the proposed algorithm takes into account the overall structure of the coefficient ma- trix and the observation vector.Conclusions:This shows the effectiveness of the proposed algorithm. The STLS estimation and the WTLS estimation ensure the statistical optimality from the perspective of the function model and the stochastic model, respectively. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Beijing University of Technology is the property of Journal of Beijing University of Technology, Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Dalian University of Technology / Dalian Ligong Daxue Xuebao is the property of Journal of Dalian University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Among the existing chaotic image encryption algorithms, there are problems such as low system security, inability to resist plaintext attacks, independent of encrypted images, and low efficiency. This paper proposed a double-graph association encryption algorithm based on a hyper-chaotic system. At first, it used a hyper-chaotic system together with the secret key to generate the key chaotic random matrix for image encryption. Then it masked plain images by the diffusion process. After that, it selected the first image to perform operations of plaintext-related scrambling and fast bit diffusion to generate cipher image. Finally, it subjected the remaining images to an adaptive correlation bit operation to generate a cipher image. Simulate results and comparative analysis show that the proposed image cryptosystem has the characteristics of large key space, the faster encryption/ decryption speed, the higher sensitivity, the better statistical properties of cipher image, and etc. And the corresponding plain text image cannot be obtained by only one cipher image. Therefore, the application prospect is greater. [ABSTRACT FROM AUTHOR]

Iterative quantization hashing algorithm does not consider that most of high-dimensional visual descriptors for images exhibit a natural matrix structure. When high-dimensional feature vectors represented the visual descriptors and assigned long binary codes,a random projection matrix requires expensive complexities in both space and time. To address the above issues, this paper proposed a hashing image retrieval method based on bilinear iterative quantization, which mapped the high-dimensional into two smaller projection matrices using compact bilinear projections instead of a single large projection matrix. Then it minimized the quantization error in an iterative quantization way and generate effective hash code. Experiments on two datasets, CIFAR-10 and Caltech256. This method achieved comparahle retrieval accuracy to the state-of-the-art 8 hashing methods while having orders of magnitude faster linear scanning time and smaller memory footprint. The results show that the proposed method can reduce the impact of the high dimensionality of data and improve the performance of ITQ, the algorithm can widely serve the hashing image retrieval application of high-dimensional data and long coded bits. [ABSTRACT FROM AUTHOR]

Copyright of Chinese Journal of Liquid Crystal & Displays is the property of Chinese Journal of Liquid Crystal & Displays and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Although it performs well in many applications, Linear discriminant analysis(LDA)is impractical for high-dimensional datasets with missing observations. One of the reasons for it is that most of the classification methods cannot predict or impute the missing values correctly, the other reason is that the sample covariance matrix used in LDA is no longer a good estimator of the population covariance matrix in high dimensions. Therefore, there will be a relatively large deviation for the discriminant function values. Based on the results from random matrix theory and by exploiting a Lasso estimator of the population covariance matrix, an improved LDA classifier for high-dimensional dataset with missing observations is proposed. Simulation results show that our proposed method is superior to the competitors for a wide variety of synthetic and real data sets. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Harbin Institute of Technology. Social Sciences Edition / Haerbin Gongye Daxue Xuebao. Shehui Kexue Ban is the property of Harbin Institute of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In the autoregressive (AR) model, random errors in the observation vector are homologous to those in the coefficient matrix. In view of the unreasonable distribution of the observation weight matrix and the inaccuracy of the random model, the random quantities in the augmented matrix consisting of the coeffi- cient matrix and the observation vector are extracted by the variable projection method. Then, we trans- form the errors-in-variables (EIV) model into the nonlinear Gauss-Helmert (GH) model and propose a struc- tural total least squares (STLS) algorithm by the nonlinear least squares adjustment theory. Combined with the least squares variance component estimation (LS - VCE) method, the variance component estimation method of STLS problem is derived. Furthermore, it is applied to the variance component estimation of the AR model. Through the real example, the effectiveness of proposed algorithm is verified. Meanwhile, the results are consistent with those of modified existing variance component estimation methods, but the construction of observation weight matrix is simple, it can also applied to the estimation of covariance factors. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Image & Graphics is the property of Editorial Office of Journal of Image & Graphics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

A hybrid cycle algorithm was proposed aiming at solving the problems that single algorithm cannot realize intelligent assembly sequence planning of complex assembly and the lack of deep human-computer interaction causes poor improvement effects. This algorithm， based on genetic algorithm， uses interference matrix and contact matrix to adjust random assembly sequence and constructs fitness function on the basis of the uniformity of assembly direction and tool. In addition， combining with simulated annealing algorithm， adding annealing operation and using Metropolis criterion， individual sequences were obtained and accepted by crossover and mutation. Meanwhile， by introducing particle swarm optimization， optimal sequence of individuals and groups were selected to cross with the offspring directly. Finally， assembly simulation platform combined with virtual reality were built and the sequence was optimized from the two dimensions of assembly stability and tool operation space. In conclusion， taking automobile rear axle assembly as an example， it is shown that the assembly sequence obtained fits actual production and the method is effective and practical. [ABSTRACT FROM AUTHOR]

In traditional Hill encryption algorithm, the modulo p multiplication of the invertible matrix and plaintext vector in Galois field GF(p) is used to calculate ciphertext vector, this paper proposed a new Hill encryption derivative algorithm in Galois field GF(2)[x]/p(x),which took polynomial high-matrix as key matrix. In this new Hill encryption derivative algorithm, it composed plaintext vector of polynomial derived from corresponding plain text, it selected a column of key matrix as translation increment randomly modulo eighth degree irreducible polynomial p(x) multiplication of the polynomial high-matrix and plaintext vector in Galois field GF(2)[x]/p(x) was done, then carried out modulo eighth degree irreducible polynomial p(x) addition of the product and translation increment in Galois field GF(2)[x]/p(x), thus it obtained the polynomial ciphertext vector, and achieved the purpose of encrypting the plaintext messages. Because it is more difficult to get plaintext from ciphertext under the condition that p(x), key matrix and random selected translation vector a re unknown, the new Hill encryption derivative algorithm in Galois field GF(2)[x]/p(x) improve the capability for anti-attack. [ABSTRACT FROM AUTHOR]

Aiming at the problem of poor precision in the application of compressed sensing theory in the field of broadband spectrum sensing, and according to the sparse characteristics of the stationary signal in the frequency domain, this paper developed a broadband compression spectrum sensing method based on P-IFourier ( partial-inverse Fourier) observation matrix. The new method translated the spectrum sensing problem into a typical compressed sensing problem, and used the standard orthogonal Fourier basis observation matrix which has the excellent incoherence performance to build the observation matrix in order to have good reconstruction performance and reconstruction precision. The simulation results show that compared with the Gaussian random observation matrix and the embedded chaotic sequence-cyclic Toeplitz structure observation matrix, this method can significantly reduce the mean square error of signal reconstruction in the lower SNR environment, and under the same conditions it can significantly improve the probability of reconstruction. [ABSTRACT FROM AUTHOR]

Aiming at the differential power attack of the encryption scheme based on lattice, Reparaz proposed an additively homomorphic R-LWE masking scheme in PQC 2016. This scheme could against the differential power attack effectively, but the additively homomorphic algorithm between the ciphertexts made the size of noise increase. It declined the rate of decryption correctness. In view of this problem, this paper proposed an improved R-LWE homomorphic masking scheme. By introducing the modular switching technology, it made a modular reduction with the cipher text of additively homomorphic. Under the premising that plaintext and the ciphertext were corresponding, it declined the size of noise in the ciphertext. And it could improve the decryption correctness of the scheme. In order to protect the sub keys, it introduced a random matrix to mask the sub key. And it made the correctness analysis and the safety proof. The proposed scheme has better security and efficiency compared to the original scheme. [ABSTRACT FROM AUTHOR]

According to the problem that the traditional information theoretic criteria has a performance decline in the case of large number of array sensors but relatively small number of sample snapshots, a novel MDL algorithm for source enumeration based on random matrix theory is proposed. The method combines some features on the distribution of eigenvalues in random matrix theory with the density of the observed data to construct a new MDL criterion based on the traditional MDL criterion framework, and then it was used for source enumeration. Simulation results and theoretical analysis show that it is consistent. Regardless of the number of array sensors, it has a high detection probability in the case of small sample snapshots number. As a result, it has a wide range of applications. Furthermore, it has a comparable computational cost compared with the traditional MDL method. [ABSTRACT FROM AUTHOR]

A method which analyses the vibrational response of the random mass plate is presented based on the Rayleigh-Ritz energy method and Lagrange dynamic equation, the response analysis method is described and the statistical characters are pointed Out. Firstly, the plate is regarded as isotropic and uniform one, as all kinds of the uncertainties of the plate can be treated with lumped mass which are subsidiary to the plate, the Rayleigh-Ritz energy method is used to analyse the vibrational energy. The energy of the lumped mass, which is random distribution on the plate, is counted, the motion equation of the variable mass plate is obtained using Lagrange dynamic equation; by solving this equation the vibrational response of the random mass plate are obtained. Appropriate Ritz base can be selected according to boundary conditions and analytical frequency bandwidth, which can improve the analytical precision of vibrational responses. It is demonstrated that the proposed method is feasible and realistic by Monte Carlo number simulation. The characteristic frequency space calculated by the presented method is the same as that predicted by random matrix theory. [ABSTRACT FROM AUTHOR]