A multi-objective memetic algorithm for low rank and sparse matrix decomposition

Wu, Tao; Shi, Jiao; Jiang, Xiangming; Zhou, Deyun; Gong, Maoguo

doi:10.1016/j.ins.2018.08.037

Cited by 12 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where C is independent of X. According to equations ( 7) and (30), equation ( 27) can be rewritten as equation (31) [35]…”

Section: Model Solvingmentioning

confidence: 99%

“…However, the failure impulses of rolling bearings are concentrated on a few large singular subspaces [27], and although penalizing all singular values can achieve noise reduction and induction of LR, it may also destroy the structure of the failure impulses. (3) The regularization parameter used to weigh the LR and sparsity in the SLR model is difficult to determine because we cannot obtain specific prior knowledge about LR and sparsity in advance [31].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Low-rank and periodic group sparse based signal denoising method for rolling bearing fault feature extraction

Zhang

Tang

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Vibration signal analysis is a vital method of achieving rolling bearing fault feature extraction, but the fault impulses contained in the vibration signals are susceptible to interference from noise, making it difficult to extract fault features. In order to effectively extract the fault features of rolling bearings, an adaptive low-rank and periodic group sparse (AdaLRPGS) denoising method is proposed. Firstly, an AdaLRPGS model is constructed, which is not only an improvement of the classical sparse low-rank method, but also can effectively enhance the low-rank and periodic group sparsity of the failure impulses. Secondly, an adaptive period prior matching method is proposed to adaptively match the period of the fault impulses, and the moth flame optimization (MFO) algorithm is utilized to adaptively search the regularization parameter of the AdaLRPGS model, and then the rule for determining the regularization parameter is summarized in the simulation analysis, which solves the dependence of the AdaLRPGS model on the period prior and regularization parameter. Finally, the solution procedure of the AdaLRPGS model is derived under the framework of the alternating direction method of multipliers (ADMM). The simulated and measured signals are analyzed using the proposed method and compared with some advanced methods. The results show that the proposed method can extract the fault features of rolling bearings and has significant advantages compared with some advanced methods.

show abstract

“…where C is independent of X. According to equations ( 7) and (30), equation ( 27) can be rewritten as equation (31) [35]…”

Section: Model Solvingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Low-rank and periodic group sparse based signal denoising method for rolling bearing fault feature extraction

Zhang

Tang

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Self-pace learning can also be combined with multi-objective optimization [ 36 ]. Furthermore, a multi-objective matrix decomposition method is proposed in [ 37 ]. For neural network structure optimization, Lu et al have proposed NSGA-Net [ 38 ] which considers the model computational cost and accuracy as an MOP and solves it with NSGA-II.…”

Section: Background and Related Workmentioning

confidence: 99%

Evolutionary Multi-Objective One-Shot Filter Pruning for Designing Lightweight Convolutional Neural Network

Shi

Zhou

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

Deep neural networks have achieved significant development and wide applications for their amazing performance. However, their complex structure, high computation and storage resource limit their applications in mobile or embedding devices such as sensor platforms. Neural network pruning is an efficient way to design a lightweight model from a well-trained complex deep neural network. In this paper, we propose an evolutionary multi-objective one-shot filter pruning method for designing a lightweight convolutional neural network. Firstly, unlike some famous iterative pruning methods, a one-shot pruning framework only needs to perform filter pruning and model fine-tuning once. Moreover, we built a constraint multi-objective filter pruning problem in which two objectives represent the filter pruning ratio and the accuracy of the pruned convolutional neural network, respectively. A non-dominated sorting-based evolutionary multi-objective algorithm was used to solve the filter pruning problem, and it provides a set of Pareto solutions which consists of a series of different trade-off pruned models. Finally, some models are uniformly selected from the set of Pareto solutions to be fine-tuned as the output of our method. The effectiveness of our method was demonstrated in experimental studies on four designed models, LeNet and AlexNet. Our method can prune over 85%, 82%, 75%, 65%, 91% and 68% filters with little accuracy loss on four designed models, LeNet and AlexNet, respectively.

show abstract

“…In the literature, one may find works that consider the multi-objective optimization to deal with signal processing problems in specific situations. For instance, Wu et al (2018) applied this approach in low rank and sparse matrix decomposition. In the context of BSS, Phlypo et al (2006) and Goh et al (2016) consider the multi-objective optimization in electroencephalogram (EEG) signal processing.…”

Section: Introductionmentioning

confidence: 99%

Application of multi-objective optimization to blind source separation

Pelegrina

Attux

Duarte

2019

Expert Systems with Applications

View full text Add to dashboard Cite

Several problems in signal processing are addressed by expert systems which take into account a set of priors on the sought signals and systems. For instance, blind source separation is often tackled by means of a mono-objective formulation which relies on a separation criterion associated with a given property of the sought signals (sources). However, in many practical situations, there are more than one property to be exploited and, as a consequence, a set of separation criteria may be used to recover the original signals. In this context, this paper addresses the separation problem by means of an approach based on multi-objective optimization. Differently from the existing methods, which provide only one estimate for the original signals, our proposal leads to a set of solutions that can be utilized by the system user to take his/her decision. Results obtained through numerical experiments over a set of biomedical signals highlight the viability of the proposed approach, which provides estimations closer to the mean squared error solutions compared to the ones achieved via a monoobjective formulation. Moreover, since our proposal is quite general, this work also contributes to encourage future researches to develop expert systems that exploit the multi-objective formulation in different source separation problems.the original sources, which are unknown in real situations. We detail this procedure in Section 4.4.2 It is worth noting that we consider in this paper the determined case, i.e., M = N . 3 In this work, we consider the minimization of cost functions. However, if J(·) must be maximized, the resulting problem can be transformed into minimization by applying the simple transformation max J(·) = min −J(·).

show abstract

A multi-objective memetic algorithm for low rank and sparse matrix decomposition

Cited by 12 publications

References 33 publications

Low-rank and periodic group sparse based signal denoising method for rolling bearing fault feature extraction

Low-rank and periodic group sparse based signal denoising method for rolling bearing fault feature extraction

Evolutionary Multi-Objective One-Shot Filter Pruning for Designing Lightweight Convolutional Neural Network

Application of multi-objective optimization to blind source separation

Contact Info

Product

Resources

About