Local Tangent Space Alignment Research Articles

Robust Hashing With Local Tangent Space Alignment for Image Copy Detection

Robust Hashing With Local Tangent Space Alignment for Image Copy Detection

Unsupervised separation of nonlinearly mixed event-related potentials using manifold clustering and non-negative matrix factorization

Event-related potentials (ERPs) can quantify brain responses to reveal the neural mechanisms of sensory perception. However, ERPs often reflect nonlinear mixture responses to multiple sources of sensory stimuli, and an accurate separation of the response to each stimulus remains a challenge. This study aimed to separate the ERP into nonlinearly mixed source components specific to individual stimuli. We developed an unsupervised learning method based on clustering of manifold structures of mixture signals combined with channel optimization for signal source reconstruction using non-negative matrix factorization (NMF). Specifically, we first implemented manifold learning based on Local Tangent Space Alignment (LTSA) to extract the spatial manifold structure of multi-resolution sub-signals separated via wavelet packet transform. We then used fuzzy entropy to extract the dynamical process of the manifold structures and performed a k-means clustering to separate different sources. Lastly, we used NMF to obtain the optimal contributions of multiple channels to ensure accurate source reconstructions. We evaluated our developed approach using a simulated ERP dataset with known ground truth of two components of ERP mixture signals. Our results show that the correlation coefficient between the reconstructed source signal and the true source signal was 92.8 % and that the separation accuracy in ERP amplitude was 91.6 %. The results show that our unsupervised separation approach can accurately separate ERP signals from nonlinear mixture source components. The outcomes provide a promising way to isolate brain responses to multiple stimulus sources during multisensory perception.

A Study on Dimensionality Reduction and Parameters for Hyperspectral Imagery Based on Manifold Learning.

With the rapid advancement of remote-sensing technology, the spectral information obtained from hyperspectral remote-sensing imagery has become increasingly rich, facilitating detailed spectral analysis of Earth's surface objects. However, the abundance of spectral information presents certain challenges for data processing, such as the "curse of dimensionality" leading to the "Hughes phenomenon", "strong correlation" due to high resolution, and "nonlinear characteristics" caused by varying surface reflectances. Consequently, dimensionality reduction of hyperspectral data emerges as a critical task. This paper begins by elucidating the principles and processes of hyperspectral image dimensionality reduction based on manifold theory and learning methods, in light of the nonlinear structures and features present in hyperspectral remote-sensing data, and formulates a dimensionality reduction process based on manifold learning. Subsequently, this study explores the capabilities of feature extraction and low-dimensional embedding for hyperspectral imagery using manifold learning approaches, including principal components analysis (PCA), multidimensional scaling (MDS), and linear discriminant analysis (LDA) for linear methods; and isometric mapping (Isomap), locally linear embedding (LLE), Laplacian eigenmaps (LE), Hessian locally linear embedding (HLLE), local tangent space alignment (LTSA), and maximum variance unfolding (MVU) for nonlinear methods, based on the Indian Pines hyperspectral dataset and Pavia University dataset. Furthermore, the paper investigates the optimal neighborhood computation time and overall algorithm runtime for feature extraction in hyperspectral imagery, varying by the choice of neighborhood k and intrinsic dimensionality d values across different manifold learning methods. Based on the outcomes of feature extraction, the study examines the classification experiments of various manifold learning methods, comparing and analyzing the variations in classification accuracy and Kappa coefficient with different selections of neighborhood k and intrinsic dimensionality d values. Building on this, the impact of selecting different bandwidths t for the Gaussian kernel in the LE method and different Lagrange multipliers λ for the MVU method on classification accuracy, given varying choices of neighborhood k and intrinsic dimensionality d, is explored. Through these experiments, the paper investigates the capability and effectiveness of different manifold learning methods in feature extraction and dimensionality reduction within hyperspectral imagery, as influenced by the selection of neighborhood k and intrinsic dimensionality d values, identifying the optimal neighborhood k and intrinsic dimensionality d value for each method. A comparison of classification accuracies reveals that the LTSA method yields superior classification results compared to other manifold learning approaches. The study demonstrates the advantages of manifold learning methods in processing hyperspectral image data, providing an experimental reference for subsequent research on hyperspectral image dimensionality reduction using manifold learning methods.

An improved transfer learning approach based on geodesic flow kernel for multiphase batch process soft sensor modeling

For multiphase batch process, the characteristics of process data under various batches differ. Consequently, the soft sensor model built for a particular working condition is inapplicable to other working conditions. Besides, each batch can be divided into several phases whose characteristics are probably different. To address the above problems, a soft sensor model based on phase division and transfer learning strategy is proposed. First, transfer learning strategy is adopted to construct a soft sensor model applicable to various working conditions. Specifically, geodesic flow kernel based on linear local tangent space alignment (LLTSA-GFK) algorithm is designed. By projecting process data to the common manifold subspace, the distribution difference of data between various batches is reduced and the performance of the soft sensor model is enhanced. In addition, sequence-based fuzzy clustering and just-in-time learning (JITL) are adopted to solve the multistage characteristic for batch process. The root-mean-square error ( RMSE), coefficient of determination [Formula: see text], and mean absolute error ( MAE) are adopted to compare the conventional soft sensing approach (i.e., partial least-square regression based on JITL, support vector regression, and back propagation neural network) with the proposed approach. The superiority of the proposed model is verified by a fed-batch penicillin fermentation process.

Improved Multi-Distance ARMF Integrated With LTSA Based Pattern Matching Method and Its Application in Fault Diagnosis

Effective dimensionality reduction (DR) and classification in fault diagnosis remain a significant challenge, primarily due to the increasing scale of industrial processes and the non-linear and high-dimensional features of process data. To address this challenge, we present local tangent space alignment (LTSA) integrated with a multi-distance adaptive order morphological filter (MARMF) fault diagnosis method (LTSA-MARMF). In LTSA-MARMF, LTSA that preserves the local manifold structure using tangent space is first utilized for DR to provide the required feature space data for ARMF. Next, the cosine distance and dynamic time warping distance are introduced into the distance error of ARMF, considering the spatial similarity and dynamic features to improve classification accuracy. Finally, the distance-matching result of the pattern is applied to determine the type of fault. Through simulations, it is evident that LTSA-MARMF can achieve more satisfactory fault diagnosis accuracy than other related methods on the Tennessee-Eastman process (TEP) and the actual Grid-connected PV System (GPVS). <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This paper is inspired by the difficult-to-handle high-dimensional and non-linear features of process data but is also applicable to high-dimensional and non-linear data from other industrial processes. The DR and classification are important aspects of fault diagnosis. In this paper, a novel pattern-matching method is utilized for fault diagnosis, which uses LTSA and the modified multi-distance ARMF for DR and classification, respectively. In terms of mathematics, the distance error of ARMF is analyzed. The combination of multi-distance is used to enhance the accuracy of fault diagnosis. The preliminary experiments show that LTSA-MARMF is feasible, but has not been tested in the plant. We will consider testing LTSA-MARMF in an actual plant in future research.

Establishing a dynamic and static knowledge model of the manufacturing cell management system: An active push approach

The workshop operation management (WOM) system contains a large amount of multi-source heterogeneous data, but these data often lack effective correlation, posing challenges in extracting valuable information for achieving unified system modelling at the knowledge level. Due to the lack of knowledge active support in the actual WOM system, managers could only passively rely on their own past experiences. To bridge this gap, the knowledge active push method for optimal operation management of the manufacturing cell (OOMMC) is studied based on knowledge graph (KG). Firstly, a dynamic knowledge model is established according to the knowledge application scenario of manufacturing cell, in which the Local Tangent Space Alignment (LTSA) + k-means algorithm is used for active abnormal warning. Secondly, a static knowledge model based on user needs using feature vocabulary and naive Bayesian classifier is established, where managers can obtain knowledge from the KG based on professional noun keywords or simple query questions. These two models complement and promote each other: the dynamic knowledge model using a service-based collaborative filtering method to push the nearest knowledges in the KG to users, effectively inspiring them to retrieve more relevant knowledge. While the static knowledge model evaluates the effectiveness of previously pushed knowledge by setting the knowledge demand index and providing feedback to the system to ensure accuracy of knowledge push. Finally, an active push simulation system is designed and developed to verify the feasibility and effectiveness of the proposed architecture and key enabling technologies.

Stacking integrated learning model via ELM and GRU with mixture correntropy loss for robust state of health estimation of lithium-ion batteries

Accurate estimation of the state of health (SOH) is crucial for the safe and stable operation of lithium-ion batteries. However, since the labeled values may contain non-Gaussian noise which may lead to a decrease in the effectiveness of traditional data-driven based estimation methods. Therefore, a novel robust stacking integrated learning model (ILM) is proposed to enable accurate SOH estimation under non-Gaussian noises (or outliers) conditions. The mixture correntropy loss (MCL) is used in original extreme learning machine (ELM) and the gated recurrent unit (GRU) frameworks to develop novel robust learning models, i.e MCL-ELM and MCL-GRU, and they are used as the sub-models of the proposed ILM, which takes into account the generalization performance of the model and the overall trend of SOH over the time series. In addition, the entropy weighting method that can well reflect the differences between the sub-models is utilized to reduce the complexity of the stacked ILM. Furthermore, the local tangent space alignment is used to capture the local relationships between different loops to enhance the effectiveness of the extracted health features. Several numerical experiments are performed under different cases to validate the estimation effect of the proposed model by using two publicly available datasets, and the experimental results demonstrate that the maximum RMSE, MAE, and MAPE values of the proposed model are 1.391%, 0.932%, and 1.480%, respectively, under non-Gaussian noises conditions.

Efficient strategies for finger movement classification using surface electromyogram signals.

One of the famous research areas in biomedical engineering and pattern recognition is finger movement classification. For hand and finger gesture recognition, the most widely used signals are the surface electromyogram (sEMG) signals. With the help of sEMG signals, four proposed techniques of finger movement classification are presented in this work. The first technique proposed is a dynamic graph construction and graph entropy-based classification of sEMG signals. The second technique proposed encompasses the ideas of dimensionality reduction utilizing local tangent space alignment (LTSA) and local linear co-ordination (LLC) with evolutionary algorithms (EA), Bayesian belief networks (BBN), extreme learning machines (ELM), and a hybrid model called EA-BBN-ELM was developed for the classification of sEMG signals. The third technique proposed utilizes the ideas of differential entropy (DE), higher-order fuzzy cognitive maps (HFCM), empirical wavelet transformation (EWT), and another hybrid model with DE-FCM-EWT and machine learning classifiers was developed for the classification of sEMG signals. The fourth technique proposed uses the ideas of local mean decomposition (LMD) and fuzzy C-means clustering along with a combined kernel least squares support vector machine (LS-SVM) classifier. The best classification accuracy results (of 98.5%) were obtained using the LMD-fuzzy C-means clustering technique classified with a combined kernel LS-SVM model. The second-best classification accuracy (of 98.21%) was obtained using the DE-FCM-EWT hybrid model with SVM classifier. The third best classification accuracy (of 97.57%) was obtained using the LTSA-based EA-BBN-ELM model.

Rapid detection of lignin content in corn straw based on Laplacian Eigenmaps

Lignin is an essential components of corn stalk and has a wide range of application. To realize the rapid detection of lignin content in corn straw and increase the detection accuracy, we initially preprocess the gathered corn straw near-infrared spectral data with Standard Normal Variable transformation (SNV). After that, the nonlinear dimensionality reduction method Laplacian Eigenmaps (LE) and Local Tangent Space Alignment (LTSA) are applied independently to reduce the dimension of spectral data. principal component analysis (PCA), a linear dimensionality reduction method, is also utilized for spectral data dimensionality reduction. Finally, models for Partial Least Squares Regression (PLSR) and Support Vector Regression (SVR) are constructed. According to the model findings, LE-SVR model offers the best prediction accuracy and stability. The determination coefficient and root mean square error of the training and tests sets are 97.17%, 0.1875 and 96.25%, 0.2718 respectively. Furthermore, the relative analytical error is 5.0776. In addition, the study findings show that the number of neighbor points k has no discernible effect on the model performance. According to the findings, nonlinear modeling using LE-SVR for NIR spectral data of corn stover can lower model complexity, while improving model prediction accuracy and stability. NIR spectroscopy may be used to determine lignin content in corn straw. At the same time, the technique in this work offers a novel approach for the rapidly detecting lignin content in other crops straw.

Voice disorder classification using convolutional neural network based on deep transfer learning

Voice disorders are very common in the global population. Many researchers have conducted research on the identification and classification of voice disorders based on machine learning. As a data-driven algorithm, machine learning requires a large number of samples for training. However, due to the sensitivity and particularity of medical data, it is difficult to obtain sufficient samples for model learning. To address this challenge, this paper proposes a pretrained OpenL3-SVM transfer learning framework for the automatic recognition of multi-class voice disorders. The framework combines a pre-trained convolutional neural network, OpenL3, and a support vector machine (SVM) classifier. The Mel spectrum of the given voice signal is first extracted and then input into the OpenL3 network to obtain high-level feature embedding. Considering the effects of redundant and negative high-dimensional features, model overfitting easily occurs. Therefore, linear local tangent space alignment (LLTSA) is used for feature dimension reduction. Finally, the obtained dimensionality reduction features are used to train the SVM for voice disorder classification. Fivefold cross-validation is used to verify the classification performance of the OpenL3-SVM. The experimental results show that OpenL3-SVM can effectively classify voice disorders automatically, and its performance exceeds that of the existing methods. With continuous improvements in research, it is expected to be considered as auxiliary diagnostic tool for physicians in the future.

Linear local tangent space alignment with autoencoder

Linear local tangent space alignment (LLTSA) is a classical dimensionality reduction method based on manifold. However, LLTSA and all its variants only consider the one-way mapping from high-dimensional space to low-dimensional space. The projected low-dimensional data may not accurately and effectively “represent” the original samples. This paper proposes a novel LLTSA method based on the linear autoencoder called LLTSA-AE (LLTSA with Autoencoder). The proposed LLTSA-AE is divided into two stages. The conventional process of LLTSA is viewed as the encoding stage, and the additional and important decoding stage is used to reconstruct the original data. Thus, LLTSA-AE makes the low-dimensional embedding data “represent” the original data more accurately and effectively. LLTSA-AE gets the recognition rates of 85.10, 67.45, 75.40 and 86.67% on handwritten Alphadigits, FERET, Georgia Tech. and Yale datasets, which are 9.4, 14.03, 7.35 and 12.39% higher than that of the original LLTSA respectively. Compared with some improved methods of LLTSA, it also obtains better performance. For example, on Handwritten Alphadigits dataset, compared with ALLTSA, OLLTSA, PLLTSA and WLLTSA, the recognition rates of LLTSA-AE are improved by 4.77, 3.96, 7.8 and 8.6% respectively. It shows that LLTSA-AE is an effective dimensionality reduction method.

Temperature field prediction of lithium-ion batteries using improved local tangent space alignment

The temperature field prediction of lithium-ion batteries (LIBs) plays a crucial role in the safety of electric vehicles and their lifetime. However, it is essentially a nonlinear distributed parameter system (DPS), and suitable partial differential equations are difficult to obtain. Further, the existing methods are weak in handling nonlinear sparse data. To address this problem, a nonlinear DPS modeling approach based on kernel and weighted local tangent space alignment (KWLTSA) is proposed in this paper. Firstly, the nonlinear dimensionality reduction method of KWLTSA is used to construct basic functions (BFs) by using spatiotemporal data. Secondly, the low-order temporal coefficients are obtained according to the spatial BFs. Subsequently, the coefficients are fed into a backpropagation neural network (BP-NN) with the input signals for training. Finally, the temperature distribution of LIBs is obtained by spatiotemporal reconstruction. Comparative experiments are designed to estimate the model performance using existing evaluation metrics. The root mean square error (RMSE) is 0.1000 and the R-square (R2) is 0.9972 of the temperature prediction by the proposed scheme. The experimental results indicate that the proposed method is more robust than the traditional dimensionality reduction method.

Manifold learning and Lempel-Ziv complexity-based fault severity recognition method for bearing

The health condition of bearings was related to the operation status of mechanical equipment, and there were relatively few methods to recognize the severity of bearing faults. Although the Lempel-Ziv complexity (LZ complexity) has been proven to be effective in recognizing the bearing fault severity, the LZ was affected by noise. And the frequency band selection-based methods were used for noise reduction for LZ, which might cause the loss of frequency band information, and affect the accuracy of LZ. Aiming at this issue, this paper proposed a bearing fault severity recognition method based on manifold learning. Firstly, the original signal was reconstructed to high-dimensional phase space. Then, local tangent space alignment was used to extract effective signals aliased in noise. Finally, the LZ complexity was obtained according to the signal after noise reduction. The proposed method was verified by constructed data and the real experimental dataset of bearing, respectively. The results showed that the proposed method can be effectively used in the fault severity recognition of bearing.

BLTSA: pseudotime prediction for single cells by branched local tangent space alignment.

The development of single-cell RNA sequencing (scRNA-seq) technology makes it possible to study the cellular dynamic processes such as cell cycle and cell differentiation. Due to the difficulties in generating genuine time-series scRNA-seq data, it is of great importance to computationally infer the pseudotime of the cells along differentiation trajectory based on their gene expression patterns. The existing pseudotime prediction methods often suffer from the high level noise of single-cell data, thus it is still necessary to study the single-cell trajectory inference methods. In this study, we propose a branched local tangent space alignment (BLTSA) method to infer single-cell pseudotime for multi-furcation trajectories. By assuming that single cells are sampled from a low-dimensional self-intersecting manifold, BLTSA first identifies the tip and branching cells in the trajectory based on cells' local Euclidean neighborhoods. Local coordinates within the tangent spaces are then determined by each cell's local neighborhood after clustering all the cells to different branches iteratively. The global coordinates for all the single cells are finally obtained by aligning the local coordinates based on the tangent spaces. We evaluate the performance of BLTSA on four simulation datasets and five real datasets. The experimental results show that BLTSA has obvious advantages over other comparison methods. R codes are available at https://github.com/LiminLi-xjtu/BLTSA. Supplementary data are available at Bioinformatics online.

Adaptive self-attention LSTM for RUL prediction of lithium-ion batteries

To achieve an accurate remaining useful life (RUL) prediction for lithium-ion batteries (LIBs), this study proposes an adaptive self-attention long short-term memory (SA-LSTM) prediction model. The innovations of the designed prediction model include the following. (1) It features an optimized local tangent space alignment algorithm, which allows the extraction of an indirect health indicator (HI) that can precisely describe battery degeneration from charge data. The extracted HI exhibits a high correlation with the standard capacity, thus facilitating RUL estimation. (2) By introducing a masked multi-head self-attention module into the time-series prediction model based on LSTM, critical information in the sequences is captured and the prediction performance is improved. (3) An online self-tuning mechanism for the weights and biases of neural networks is designed to correct cumulative estimation errors in long-term predictions and reduce the effects of local fluctuations and regeneration. The proposed prediction model enables the HI values in future cycles to be iteratively estimated using the one-step-ahead method, and the RUL can be forecast once the predicted signal falls. Experimental results indicate the effectiveness and superiority of the proposed prediction method.

Weighted Linear Local Tangent Space Alignment via Geometrically Inspired Weighted PCA for Fault Detection

Principal component analysis (PCA) is widely adopted in local tangent space alignment to estimate local tangent spaces. These estimates are only accurate when uniformly distributed data lies in or is close to linear subspaces. In practice, such conditions are rarely satisfied. Therefore, this approach fails to reveal manifold intrinsic features, resulting in degraded fault detection accuracy. Considering the drawbacks, weighted linear local tangent space alignment (WLLTSA), a manifold learning method is put forward. First, weighted PCA is adopted to provide local tangent space estimates. The parameter selection criterion for the weight matrix is established by taking the context of geometric preservation into account. Second, global low dimensional coordinates are formed by aligning local coordinates with global feature space. Finally, the fault detection model is developed, and kernel density estimation is utilized to approximate confidence bounds for <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathrm{T}^{\mathrm{2}}$</tex-math></inline-formula> and SPE statistics. Simulation results are presented to illustrate the superior feature extraction and fault detection performance of WLLTSA.

Adaptive Fault Diagnosis and Decision-Making Method Based on Multi-Spectrum Evaluation and Fusion for Traction Motor Bearings

Traction motor bearings are critical research objects of motor fault diagnosis due to their high failure rate. Nowadays, numerous frequency-domain analysis methods have been proposed for rolling bearing fault diagnosis. However, most methods have rigorous requirements on the applicable conditions and may fail to diagnose when working conditions change, which significantly limits the broad application of the method in practical engineering. Therefore, this article presents an adaptive fault diagnosis and decision-making method (AFDDM) based on multi-spectrum evaluation and fusion. The method can give appropriate weights to each spectrum in different conditions to ensure the stability of fused spectrum and convert the spectrums into more intuitive fault probability to give maintenance advice. Also, the AFDDM method achieves the purpose of evaluating the spectrum by extracting the eigenvalue vector and utilizing the local tangent space alignment (LTSA) to reduce dimension. Then, a decision-making model based on multi-spectrum is proposed to divert the spectrums into fault probability. By this method, the evaluation fusion of spectrums with different effects and decision-making is realized. Real bearing fault signals from different data sets are selected to verify the effectiveness of the proposed method and show that the proposed method has better engineering applicability and robustness.

Fault diagnosis model based on adaptive generalised morphological filtering and LLTSA-ELM

It is difficult for a single feature to contain all the information needed to describe the running state of the equipment. Though multi-features can contain more information about running state, the redundancy between high-dimension features can easily reduce the accuracy of the classifier. Aimed at that, a fault diagnosis method for rolling bearings combining adaptive generalised morphological filter, linear local tangent space alignment and extreme learning machine (LLTSA-ELM) is proposed. Firstly, the rolling bearing vibration signals are filtered by an adaptive generalised morphological filter. Secondly, the multi-domain features are extracted from filtered signal to construct high-dimensional features set of bearing. Thirdly, the dimension of high-dimensional features is reduced by maximum likelihood estimation (MLE) and LLTSA. Finally, the bearing condition monitoring model based on ELM is constructed by the reduced dimension features, and then is used to analyse and diagnose the running state of bearing. Experiments show that this method can effectively filter noise, reduce the redundancy caused by high-dimensional features, and improve the recognition accuracy of the running state of bearings.

Cross-Subject Emotion Recognition Based on Domain Similarity of EEG Signal Transfer Learning.

For solving the problem of the inevitable decline in the accuracy of cross-subject emotion recognition via Electroencephalograph (EEG) signal transfer learning due to the negative transfer of data in the source domain, this paper offers a new method to dynamically select the data suitable for transfer learning and eliminate the data that may lead to negative transfer. The method which is called cross-subject source domain selection (CSDS) consists of the next three parts. 1) First, a Frank-copula model is established according to Copula function theory to study the correlation between the source domain and the target domain, which is described by the Kendall correlation coefficient. 2) The calculation method for the Maximum Mean Discrepancy is improved to determine the distance between classes in a single source. After normalization, the Kendall correlation coefficient is superimposed, and the threshold is set to identify the source-domain data most suitable for transfer learning. 3) In the process of transfer learning, on the basis of Manifold Embedded Distribution Alignment, the Local Tangent Space Alignment method is used to provide a low-dimensional linear estimation of the local geometry of nonlinear manifolds, which maintains the local characteristics of the sample data after dimensionality reduction. Experimental results show that compared with the traditional methods, the CSDS increases the accuracy of emotion classification by approximately 2.8% and reduces the runtime by approximately 65%.

Application of multi-angle millimeter-wave radar detection in human motion behavior and micro-action recognition

Millimeter-wave radar is widely used in family safety, rehabilitation, and assisted living due to its ability to operate in all weathers and all day. To address the problem whereby the radar detection angle significantly impacts human behavior recognition, a recognition method based on multi-angle radar observation is adopted. We proposed a novel radar selection method called the energy domain ratio method to choose a radar with more sensitive features. Then, local tangent space alignment and an adaptive extreme learning machine (ELM) are presented to enhance the recognition rate of the model in a high-noise environment. A multi-angle entropy feature and an improved ELM are developed to identify human micro-motion in a low-noise indoor environment. The effect of observation distance on the recognition effect was also explored. The experimental results show that the proposed model has a more than 86% recognition rate for human behavior in outdoor scenes and a recognition accuracy of more than 98% for indoor micro-action.