Abstract The prevalence of cardiovascular disease (CVD) has surged in recent years, making it the foremost cause mortality among humans. Electrocardiogram (ECG), being one pivotal diagnostic tools for diseases, is increasingly gaining prominence field machine learning. However, prevailing neural network models frequently disregard spatial dimension features inherent ECG signals. In this paper, we propose an autoencoder architecture incorporating low-rank attention (LRA-autoencoder). It...

Depression diagnosis is easily affected by subjective consciousness.It of great significance to study objective and accurate identification methods. Electroencephalogram (EEG) can reflect brain activity working state. Therefore, this paper aims explore features with significant differences based on functional connectivity improve the accuracy depression recognition.We propose a Functional Connection Feature Selection Fuzzy Label (FLFCFS), it calculates correlation between electrode pairs...

When utilising non‐negative matrix factorisation (NMF) to decompose a data into the product of two low‐rank matrices with entries, noisy components may be introduced matrix. Many approaches have been proposed address problem. Different from them, authors consider group sparsity and geometric structure by introducing ‐norm local preserving regularisation in formulated objective function. A graph regularised sparse NMF de‐noising approach is learn discriminative representations for original...

This paper primarily explores the challenge of stealthy innovation-based attacks within cyber–physical systems. An optimal covert false data injection attack strategy is proposed, which involves modifying historical and current system innovations by adversary. The approach differs from existing ones in that it combines innovation results. Utilizing less prior knowledge to obtain made adversary previous time step integrating them with innovations, a more generalized devised, leading further...

Abstract This paper presents a new fault tolerant controller design method for class of interconnected non‐Gaussian stochastic distribution system with boundary conditions. In order to obtain the estimation value, an observer based detection and diagnosis algorithms are presented at first, then collaborative is designed on adaptive control strategy. Different from most existing controllers, when occurs need be reconstructed healthy subsystem in this paper. That say, compensated not by faulty...

Mild cognitive impairment (MCI) represents a transitional stage between normal aging and Alzheimer's disease (AD), with higher risk to convert AD. The information of AD control (NC) subjects can aid the classification progressive MCI stable MCI. In this paper, we develop an effective biomarker by combining auxiliary NC relationship brain regions subject, which makes best improves prediction accuracy MCI-to-AD conversion. Specifically, projection vector is first obtained for each subject via...

To date, quality-related multivariate statistical methods are extensively used in process monitoring and have achieved admirable effects. However, most of them contain recursive processes, which result higher time complexity not suitable for increasingly complex industrial processes. Therefore, this paper embeds singular value decomposition (SVD) into the kernel principal component regression (KPCR) to accomplish Quality-related with a lower computational cost. Specifically, technique is...

Aiming at the sequential signal transmission for an interconnected control system with a non-Gaussian random distribution, we design Adaptive Neural Network Collaborative Fault-Tolerant Controller (ANN-CFTC). ANN-CFTC adopts adaptive neural networks to directly estimate unknown couplings and faults. The collaborative fault-tolerant controller is designed by means of fault-free subsystem compensate fault. This can better guarantee stability overall system, not single subsystem. Simulation...

Accurate recognition of progressive mild cognitive impairment (MCI) is helpful to reduce the risk developing Alzheimer’s disease (AD). However, it still challenging extract effective biomarkers from multivariate brain structural magnetic resonance imaging (MRI) features accurately differentiate MCI stable MCI. We develop novel by combining subspace learning methods with information AD as well normal control (NC) subjects for prediction conversion using MRI data. Specifically, we first learn...

Abstract This paper primarily focuses on designing stealthy false data injection attacks targeting two communication channels in cyber‐physical systems equipped with state estimators and attack detectors. It introduces the concept of perfect attacks, rendering detector unable to detect designed signals, thereby enabling attackers further destabilize system. Through error model by attacker, does not require knowledge estimator. By injecting a carefully into channel, it is possible extend...

The ECG signal is often accompanied by noise, which can affect its shape characteristics, so it important to perform de-noising. However, the commonly used noise reduction methods, such as wavelet or filter transformation, prioritize high-frequency signals over low-frequency ones, leading loss of band features difficulties in capturing them. We propose a fusion reconstruction framework that combines hash autoencoder and margin semantic reinforcement enhance features. Specifically, for...