One-dimensional Input Research Articles

Barlow Twins deep neural network for advanced 1D drug–target interaction prediction

Accurate prediction of drug–target interactions is critical for advancing drug discovery. By reducing time and cost, machine learning and deep learning can accelerate this laborious discovery process. In a novel approach, BarlowDTI, we utilise the powerful Barlow Twins architecture for feature-extraction while considering the structure of the target protein. Our method achieves state-of-the-art predictive performance against multiple established benchmarks using only one-dimensional input. The use of our hybrid approach of deep learning and gradient boosting machine as the underlying predictor ensures fast and efficient predictions without the need for substantial computational resources. We also propose the use of an influence method to investigate how the model reaches its decision based on individual training samples. By comparing co-crystal structures, we find that BarlowDTI effectively exploits catalytically active and stabilising residues, highlighting the model’s ability to generalise from one-dimensional input data. In addition, we further benchmark new baselines against existing methods. Together, these innovations improve the efficiency and effectiveness of drug–target interactions predictions, providing robust tools for accelerating drug development and deepening the understanding of molecular interactions. Therefore, we provide an easy-to-use web interface that can be freely accessed at https://www.bio.nat.tum.de/oc2/barlowdti.Scientific contributionOur computationally efficient and effective hybrid approach, combining the deep learning model Barlow Twins and gradient boosting machines, outperforms state-of-the-art methods across multiple splits and benchmarks using only one-dimensional input. Furthermore, we advance the field by proposing an influence method that elucidates model decision-making, thereby providing deeper insights into molecular interactions and improving the interpretability of drug-target interactions predictions.Graphical

Single-pixel deep phase-shifting incoherent digital holography

Incoherent digital holography technology reduces the requirement for coherence of light sources, greatly expanding the application range of digital holography. In this paper, we designed a Multi-head attention single-pixel (MHASP) phase-shifting network for incoherent digital holography. The trained network has the capability to effortlessly predict three interferograms, encompassing phase shifts of 0, 2/3 π, and 4/3 π, solely from one-dimensional input data. Utilizing the conventional three-step phase-shifting method, we are able to effectively eliminate the DC and twin terms from the holographic reconstruction process, subsequently achieving a high-fidelity reconstruction facilitated by the employment of the back propagation algorithm. The comprehensive experimental findings clearly indicate that, beyond facilitating high-precision reconstruction, the introduced MHASP phase-shifting approach efficiently preserves 3D information through calibrating the back propagation distance, even when confronted with a reduced volume of holographic data. Furthermore, the introduced approach uses a network to replace the actual phase shift operation, which can better improve the utilization of object light energy. This approach not only circumvented the constraints posed by area array sensors but also facilitated high-fidelity imaging with minimal data volume, thereby expanding the horizons of incoherent digital holography applications in the realm of 3D imaging.

Shaking table tests and numerical analysis of monopile-supported offshore wind turbines under combined wind, wave and seismic loads

Shaking table tests and numerical analysis of monopile-supported offshore wind turbines under combined wind, wave and seismic loads

Forward modeling of quake’s infrasound recorded in the stratosphere on board balloon platforms

Acoustic waves generated by seismic waves contain information on the internal structure of planets, and can be sensed by pressure sensors onboard high-altitude balloons. To identify the various contributions (infrasound signal, noise, balloon response, etc.) in such pressure records, a full waveform modeling is implemented and completed by infrasound ray tracing and additional data analysis. Here, we analyze the Stratéole-2 pressure data associated with two earthquakes (Garcia et al. Geophys. Res. Lett. 49(15):e98844, 2022) and compared these to full waveform simulations by SPECFEM2D-DG-LNS software. Even if our simulations do not precisely reproduce the waveform observed in the frequency range [0.05, 0.3] Hz, we show that the waveform presents more sensitivity to quake and internal structure parameters than to atmospheric structure, and that seismic surface wave dispersion is observed in balloon pressure records. The long-duration pressure oscillations observed after the main infrasonic signal cannot be fully reproduced by our one-dimensional input model even when source time function complexity and aftershocks are considered. These features are ascribed mainly to the complex vertical ground movements below the balloon and partly to late secondary infrasound arrivals excited by the interactions of seismic waves with the topography. These results enhance the advantages and limitations of quake-related infrasound observations on board terrestrial and planetary balloon platforms.Graphical

Asymptotic Form of the Covariance Matrix of Likelihood-Based Estimator in Multidimensional Linear System Model for the Case of Infinity Number of Nuisance Parameters

This article is devoted to the synthesis and analysis of the quality of the statistical estimate of parameters of a multidimensional linear system (MLS) with one input and m outputs. A nontrivial case is investigated when the one-dimensional input signal of MLS is a deterministic process, the values of which are unknown nuisance parameters. The estimate is based only on observations of MLS output signals distorted by random Gaussian stationary m-dimensional noise with a known spectrum. It is assumed that the likelihood function of observations of the output signals of MLS satisfies the conditions of local asymptotic normality. The n-consistency of the estimate is established. Under the assumption of asymptotic normality of an objective function, the limiting covariance matrix of the estimate is calculated for case where the number of observations tends to infinity.

TransQA: deep hybrid transformer network for measurement-guided volumetric dose prediction of pre-treatment patient-specific quality assurance

Objective. Performing pre-treatment patient-specific quality assurance (prePSQA) is considered an essential, time-consuming, and resource-intensive task for volumetric modulated arc radiotherapy (VMAT) which confirms the dose accuracy and ensure patient safety. Most current machine learning and deep learning approaches stack excessive convolutional/pooling operations (CPs) to predict prePSQA with two-dimensional or one-dimensional information input. However, these models generally present limitations in explicitly modeling long-range dependency for volumetric dose prediction due to the loss of spatial dose features and the inherent locality of CPs. The purpose of this work is to construct a deep hybrid network by combining the self-attention mechanism-based Transformer with modified U-Net for predicting measurement-guided volumetric dose (MDose) of prePSQA. Approach. The enrolled 307 cancer patients underwent VMAT were randomly divided into 246 and 61 cases for training and testing the model. The input data included computed tomography images, radiotherapy dose images exported from the treatment planning system, as well as the MDose distribution from the verification system. The output was the predicted high-quality voxel-wise prePSQA dose distribution. Main results: qualitative and quantitative experimental results show that the proposed prediction method could achieve comparable or better performance on MDose prediction over other approaches in terms of spatial dose distribution, dose–volume histogram metrics, gamma passing rates, mean absolute error, root mean square error, and structural similarity. Significance. The preliminary results on multiple cancer sites show that our approach can be taken as a clinical guidance tool and help medical physicists to reduce the measurement work of prePSQA.

MITDCNN: A multi-modal input Transformer-based deep convolutional neural network for misfire signal detection in high-noise diesel engines

MITDCNN: A multi-modal input Transformer-based deep convolutional neural network for misfire signal detection in high-noise diesel engines

Key technologies research on of soil-structure interaction base story isolated structure response in 3D seismic zone

The development of karst in Karst area leads to poor stability of stratum. If earthquake occurs, the area will produce destructive disaster. In order to improve the stability capacity of the grassroots in the region, this study investigates the seismic response of inter-story isolation structures considering soil-structure interaction (SSI) in three-dimensional earthquakes. A model of the inter-story isolation structure incorporating SSI was developed, and one-dimensional, two-dimensional, and three-dimensional ground motions were applied to compare the seismic response under different input conditions. A three-dimensional isolation system was introduced and compared with traditional horizontal isolation structures to address excessive tensile and compressive stresses on the isolation structure during three-dimensional ground motion. The results demonstrate that the seismic response to three-dimensional earthquakes surpasses one-dimensional and two-dimensional inputs. Furthermore, adding a three-dimensional isolation structure effectively isolates vertical ground motion and reduces structural seismic response. Moreover, it minimizes soil stresses on the foundation compared to traditional horizontal isolation structure, enhancing foundation stability. This study will provide theoretical value and practical guidance for the research on key technology of SSI base story isolation structure response in Karst Plateau 3D Seismic zone.

2D-WinSpatt-Net: A Dual Spatial Self-Attention Vision Transformer Boosts Classification of Tetanus Severity for Patients Wearing ECG Sensors in Low- and Middle-Income Countries.

Tetanus is a life-threatening bacterial infection that is often prevalent in low- and middle-income countries (LMIC), Vietnam included. Tetanus affects the nervous system, leading to muscle stiffness and spasms. Moreover, severe tetanus is associated with autonomic nervous system (ANS) dysfunction. To ensure early detection and effective management of ANS dysfunction, patients require continuous monitoring of vital signs using bedside monitors. Wearable electrocardiogram (ECG) sensors offer a more cost-effective and user-friendly alternative to bedside monitors. Machine learning-based ECG analysis can be a valuable resource for classifying tetanus severity; however, using existing ECG signal analysis is excessively time-consuming. Due to the fixed-sized kernel filters used in traditional convolutional neural networks (CNNs), they are limited in their ability to capture global context information. In this work, we propose a 2D-WinSpatt-Net, which is a novel Vision Transformer that contains both local spatial window self-attention and global spatial self-attention mechanisms. The 2D-WinSpatt-Net boosts the classification of tetanus severity in intensive-care settings for LMIC using wearable ECG sensors. The time series imaging-continuous wavelet transforms-is transformed from a one-dimensional ECG signal and input to the proposed 2D-WinSpatt-Net. In the classification of tetanus severity levels, 2D-WinSpatt-Net surpasses state-of-the-art methods in terms of performance and accuracy. It achieves remarkable results with an F1 score of 0.88 ± 0.00, precision of 0.92 ± 0.02, recall of 0.85 ± 0.01, specificity of 0.96 ± 0.01, accuracy of 0.93 ± 0.02 and AUC of 0.90 ± 0.00.

Learning Mappings from Iced Airfoils to Aerodynamic Coefficients Using a Deep Operator Network

This study abstracted the prediction of the aerodynamic coefficients of an iced airfoil as a mapping from the iced airfoil space to the aerodynamic coefficient space. Thus, a deep network called Airfoils2AeroNet that can learn this mapping was established based on Deep Operator Network (DeepONet). The deep network consists of a branch network for encoding the iced airfoil images and a trunk network that learns a nonlinear mapping from the one-dimensional aerodynamic coefficient function input to p-dimensional outputs. The branch network consists of deep convolutional neural networks (CNNs), and the trunk network consists of fully connected neural networks (FNNs). Then the network was trained and tested on iced airfoils based on NACA 0012 airfoils. Comparing the prediction results of Airfoils2AeroNet with those of the conventional direct CNN network, the network proposed in this paper has a strong advantage for generalization. Unlike the traditional CNN, which can only predict the aerodynamic coefficients at fixed flow conditions consistent with the training data, the network can flexibly predict the aerodynamic coefficients at different flow conditions. Finally, the influence of the structure of the branch network and trunk network on the prediction results was analyzed.

A new on-line self-adapting fuzzy controller design using unidimensional input-output with dynamical membership functions

Here, we develop a fuzzy controller using a new online self-adapting design. The objective of this work is to control a nonlinear process by using a one-dimensional input rule variable, instead of error and error variation. The initial limits of the fuzzy logic membership functions are mostly depend on experiments and previous knowledge of the dynamic process behaviors. Generally, the membership function parameters have a significant impact on control signal amplitude and, consequently on the convergence and stability of the controller-plant system. The proposed technique determines the limits of the antecedent membership functions online using the kth and k - 1th outputs of the controlled plant and reference model, respectively. Meanwhile, the limits of the consequent membership functions are calculated using error and error variation. This approach ensures: (i) that the input/output variables have the required fuzzy space, (ii) the controlled plant follows the desired reference model, and (iii) the control signal amplitude is within acceptable limits. Additionally, (iiii) it takes into account the dynamic variability of the process and the existence of an overshoot. The membership function parameters are updated continuously through a self-adapting procedure, ensuring improved control performance. Ultimately, the proposed approach is improved using two nonlinear systems.

Improving Classification of Tetanus Severity for Patients in Low-Middle Income Countries Wearing ECG Sensors by Using a CNN-Transformer Network.

Tetanus is a life-threatening infectious disease, which is still common in low- and middle-income countries, including in Vietnam. This disease is characterized by muscle spasm and in severe cases is complicated by autonomic dysfunction. Ideally continuous vital sign monitoring using bedside monitors allows the prompt detection of the onset of autonomic nervous system dysfunction or avoiding rapid deterioration. Detection can be improved using heart rate variability analysis from ECG signals. Recently, characteristic ECG and heart rate variability features have been shown to be of value in classifying tetanus severity. However, conventional manual analysis of ECG is time-consuming. The traditional convolutional neural network (CNN) has limitations in extracting the global context information, due to its fixed-sized kernel filters. In this work, we propose a novel hybrid CNN-Transformer model to automatically classify tetanus severity using tetanus monitoring from low-cost wearable sensors. This model can capture the local features from the CNN and the global features from the Transformer. The time series imaging - spectrogram - is transformed from one-dimensional ECG signal and input to the proposed model. The CNN-Transformer model outperforms state-of-the-art methods in tetanus classification, achieves results with a F1 score of 0.82±0.03, precision of 0.94±0.03, recall of 0.73±0.07, specificity of 0.97±0.02, accuracy of 0.88±0.01 and AUC of 0.85±0.03. In addition, we found that Random Forest with enough manually selected features can be comparable with the proposed CNN-Transformer model.

Admissibility of retarded diagonal systems with one-dimensional input space

Admissibility of retarded diagonal systems with one-dimensional input space

A novel attention-guided ECA-CNN architecture for sEMG-based gait classification.

Gait recognition and classification technology is one of the essential technologies for detecting neurodegenerative dysfunction. This paper presents a gait classification model based on a convolutional neural network (CNN) with an efficient channel attention (ECA) module for gait detection applications using surface electromyographic (sEMG) signals. First, the sEMG sensor was used to collect the experimental sample data, and various gaits of different persons were collected to construct the sEMG signal data sets of different gaits. The CNN is used to extract the features of the one-dimensional input sEMG signal to obtain the feature vector, which is input into the ECA module to realize cross-channel interaction. Then, the next part of the convolutional layer is input to learn the signal features further. Finally, the model is output and tested to obtain the results. Comparative experiments show that the accuracy of the ECA-CNN network model can reach 97.75%.

Remarkable robustness of self-imaging distance for a misaligned paraxial Gaussian input and variation in the nonparaxial regime

The variation of focusing distance in a parabolic graded-index slab with the width of a one-dimensional Gaussian input fed at its waist, both axially and misaligned, into the waveguide is studied in paraxial and beyond-paraxial regimes. We obtain analytical expressions, scalable in terms of material parameters, for input coupling coefficients for such a Gaussian input. The focusing distance shows remarkable stability for an axially fed input for beam width exceeding the fundamental mode width of the waveguide. There is a smooth variation for the other regime of beam width. In the paraxial domain, we identify a unique beam width of ∼0.76 times the fundamental mode width for which the self-imaging distance is nearly independent of misalignment. The stability, a well-known sharp shift of the focusing point for an axially fed beam of width around that of the fundamental mode, and remarkable stability of self-imaging distance with misalignment at the unique beam width should be useful for efficiency enhancement of device interconnects, sensing, and lensing applications.

Detection of magnetohydrodynamic waves by using convolutional neural networks

Nonlinear wave interactions in magnetohydrodynamics (MHD), such as shock refraction at an inclined density interface, lead to a plethora of wave patterns with numerous wave types. Identification of different types of MHD waves is an important and challenging task in such complex wave patterns. Moreover, owing to the multiplicity of solutions and their admissibility for different systems, especially for intermediate-type MHD shock waves, the identification of MHD wave types is complicated if one relies on the Rankine–Hugoniot jump conditions. MHD wave detection is further exacerbated by nonphysical smearing of discontinuous shock waves in numerical simulations. This paper proposes two MHD wave detection methods based on convolutional neural network to enable wave classification and identify their locations. The first method separates the output into regression (location prediction) and classification problems, assuming the number of waves for each training data is fixed. In contrast, the second method does not specify the number of waves a priori, and the algorithm predicts wave locations and classifies types using only regression. We use one-dimensional input data (density, velocity, and magnetic fields) to train the two models that successfully reproduce a complex two-dimensional MHD shock refraction structure. The first fixed output model efficiently provides high precision and recall, achieving total neural network accuracy up to 99%, and the classification accuracy of some waves approaches unity. The second detection model has relatively low performance, with more sensitivity to the setting of parameters, such as the number of grid cells Ngrid and the thresholds of confidence score and class probability, etc. The detection model achieves better than 90% accuracy with F1 score >0.95. The proposed two methods demonstrate very strong potential for MHD wave detection in complex wave structures and interactions.

On the Existence of Global Minima and Convergence Analyses for Gradient Descent Methods in the Training of Deep Neural Networks

In this article we study fully-connected feedforward deep ReLU ANNs with an arbitrarily large number of hidden layers and we prove convergence of the risk of the GD optimization method with random initializations in the training of such ANNs under the assumption that the unnormalized probability density function of the probability distribution of the input data of the considered supervised learning problem is piecewise polynomial, under the assumption that the target function (describing the relationship between input data and the output data) is piecewise polynomial, and under the assumption that the risk function of the considered supervised learning problem admits at least one regular global minimum. In addition, in the special situation of shallow ANNs with just one hidden layer and one-dimensional input we also verify this assumption by proving in the training of such shallow ANNs that for every Lipschitz continuous target function there exists a global minimum in the risk landscape. Finally, in the training of deep ANNs with ReLU activation we also study solutions of gradient flow (GF) differential equations and we prove that every non-divergent GF trajectory converges with a polynomial rate of convergence to a critical point (in the sense of limiting Fr\'echet subdifferentiability). Our mathematical convergence analysis builds up on ideas from our previous article Eberle et al., on tools from real algebraic geometry such as the concept of semi-algebraic functions and generalized Kurdyka-Lojasiewicz inequalities, on tools from functional analysis such as the Arzel\`a-Ascoli theorem, on tools from nonsmooth analysis such as the concept of limiting Fr\'echet subgradients, as well as on the fact that the set of realization functions of shallow ReLU ANNs with fixed architecture forms a closed subset of the set of continuous functions revealed by Petersen et al.

A simple modelling strategy for integer order and fractional order interval type-2 fuzzy PID controllers with their simulation and real-time implementation

This paper presents a simple approach to the mathematical modelling of the Integer Order (IO) and Fractional Order (FO) Interval Type-2 Fuzzy Proportional Integral Derivative (IT2FPID) controllers. In this approach, the control effort produced by the IT2FPID controller is obtained by combining the individual control efforts generated by the Interval Type-2 Fuzzy Proportional (IT2FP), Interval Type-2 Fuzzy Integral (IT2FI), and Interval Type-2 Fuzzy Derivative (IT2FD) controllers. The analytical structures of each of the IT2FPID components are unveiled using one-dimensional (1D) input space and without using any AND or OR operator. To the best of the authors’ knowledge, such a strategy was never used and is completely new for the modelling of the IT2FPID controllers. Properties, computational issues, and design guidelines of the newly obtained controllers are discussed, and their applicability is delineated using five simulation examples and one experimental case study. A chemical reactor plant, two fractional order plants, a nonlinear plant, and a Twin Rotor Multi Input Multi Output System (TRMS) are considered in the simulation study. The experimental study is conducted on an unstable nonlinear Magnetic Levitation System (MLS). Moreover, to exhibit the usefulness of the newly derived controllers, simulation and real-time results are also compared with the existing results available in the literature. As the proposed controllers are plant model-free, they can be used for other control applications also.

Optimal designs for some bivariate cokriging models

This article focuses on the estimation and design aspects of a bivariate collocated cokriging experiment. For a large class of covariance matrices, a linear dependency criterion is identified, which allows the best linear unbiased estimator of the primary variable in a bivariate collocated cokriging setup to reduce to a univariate kriging estimator. Exact optimal designs for efficient prediction for such simple and ordinary reduced cokriging models with one-dimensional inputs are determined. Designs are found by minimizing the maximum and the integrated prediction variance, where the primary variable is an Ornstein-Uhlenbeck process. For simple and ordinary cokriging models with known covariance parameters, the equispaced design is shown to be optimal for both criterion functions. The more realistic scenario of unknown covariance parameters is addressed by assuming prior distributions on the parameter vector, thus adopting a Bayesian approach to the design problem. The equispaced design is proved to be the Bayesian optimal design for both criteria. The work is motivated by designing an optimal water monitoring system for an Indian river.

CLASSIFICATION AND DETECTION OF DYSRHYTHMIA FOR LECTROCARDIOGRAPHY SIGNALS BY CONVOLUTIONAL NEURAL NETWORK

Electrocardiography is the most useful method for diagnosing cardiovascular disease such as arrhythmia. Heartbeat classification of electrocardiography signals is a valuable and hopeful technology for early warning of dysrhythmia because it contains the cardiac electrical activities and reflects the abnormal cardiac activity. Therefore, a novel electrocardiogram-based arrhythmic beats classification was proposed to automatically detect the main types of dysrhythmia using electrocardiography signal in this work. A convolutional neural network model was used to automatically detect the normal and the types of dysrhythmia electrocardiography beats. The beat was transformed into a matrix as two-dimensional input to the model. The classification system was assessed to detect the normal, left bundle branch block, premature ventricular contraction and right bundle branch block beats using the MIT-BIH arrhythmia database. The results showed that an average accuracy of 99.30% and 98.85% was achieved by using ML2 lead of electrocardiography data with one-dimensional and 2-dimensional input, respectively. An average accuracy of 97.00% and 97.20% was achieved by using V1 lead of electrocardiography data with one-dimensional and 2-dimensional input, respectively. Moreover, no feature extraction of signals was carried out in this study. Consequently, the proposed model can accurately test the unknown electrocardiography signal and aid the clinician in the diagnosis of dysrhythmia.