Sparse Feature Extraction Research Articles

A Sparse Feature Extraction Method with Elastic Net for Drug-Target Interaction Identification

The identification of drug-target interactions (DTIs) plays a crucial role in drug discovery. However, the traditional high-throughput techniques based on clinical trials are costly, cumbersome, and time-consuming for identifying DTIs. Hence, new intelligent computational methods are urgently needed to surmount these defects in predicting DTIs. In this paper, we propose a novel computational method that combines position-specific scoring matrix (PSSM), elastic net based sparse features extraction, and rotation forest (RF) classifier. Specifically, we converted each protein primary sequence into PSSM, which contains biological evolutionary information. Then we extract the hidden sparse feature descriptors in PSSM by elastic net based sparse feature extraction method (ESFE). After that, we fuse them with the features of drug, which are represented by molecular fingerprints. Finally, rotation forest classifier works on detecting the potential drug-target interactions. When performing the proposed method by the experiments of fivefold cross validation (CV) on enzyme, ion channel, G protein-coupled receptors (GPCRs), and nuclear receptor datasets, this method achieves average accuracies of 90.32%, 88.91%, 80.65%, and 79.73%, respectively. We also compared the proposed model with the state-of-the-art support vector machine (SVM) classifier and other effective methods on the same datasets. The comparison results distinctly indicate that the proposed model possesses the efficient and robust ability to predict DTIs. We expect that the new model will be able to take effects on predicting massive DTIs.

Landslide Susceptibility Prediction Using Sparse Feature Extraction and Machine Learning Models Based on GIS and Remote Sensing

Landslide susceptibility prediction (LSP) is a useful technology for landslide prevention. Due to the complex nonlinear correlations among environmental factors, traditional machine learning (ML) models have unsatisfactory LSP accuracies. In this letter, a sparse feature extraction network (SFE+) is proposed for LSP. First, the landslides and environmental factors are collected, and frequency ratios of environmental factors are calculated as the model inputs. Second, the input data are passed through the input layer with the dropout, and then, the features are passed through the hidden layers, that is, the k% lifetime sparsity layers. The hidden layers are employed to further sparse these factors to obtain the independent and redundant prediction features as much as possible. Finally, certain classifiers are used to realize the LSP in the study area. SFE-support vector machine (SVM), SFE-logistic regression (LR), and SFE-stochastic gradient descent (SGD) models are built. For comparison, principal component analysis (PCA)-SVM, PCA-LR, PCA-SGD, SVM, LR, and SGD models are also built for LSP in Shicheng County, China. Results show that the SFE-based ML models, especially the SFE-SVM, can effectively extract the sparse nonlinear features of environmental factors to improve LSP accuracies and have promising prospects for LSP.

WLnet: Towards an Approach for Robust Workload Estimation Based on Shallow Neural Networks

Electroencephalography (EEG) is a non-invasive technology used for the human brain-computer interface. One of its important applications is the evaluation of the mental state of an individual, such as workload estimation. In previous works, common spatial pattern feature extraction methods have been proposed for the EEG-based workload detection. Recently, several novel methods were introduced to detect EEG pattern workloads. However, it is still unknown which one of these methods is the one that offers the best performance for the workload EEG pattern feature detection. In this article, four methods were used to extract workload EEG features: (a) common spatial pattern feature extraction; (b) temporally constrained sparse group spatial pattern feature extraction; (c) EEGnet; and (d) the new proposed shallow convolutional neural network for workload estimation (WLnet). The classification accuracy of these four methods was compared. Experimental results demonstrate that the proposed WLnet achieved the best detection accuracy in both stress and non-stress conditions. We believe that the proposed methods may be relevant to real-life applications of mental workload estimation.

Sparse feature extraction based on periodical convolutional sparse representation for fault detection of rotating machinery

Sparse fault transient extraction is the primary step in rotating machine fault detection. In the present paper, periodical convolutional sparse representation (PCSR) is proposed for reliable separation of fault transients imbedded in raw vibration signals. Specifically, a sparse optimization problem of PCSR is constructed, in which periodical fault transients and harmonic components are sparsely represented by a learned dictionary and Fourier dictionary, and the periodicity and group sparsity of sparse coefficients related to sparse fault transients are also incorporated. Meanwhile, to further promote the sparsity of the sparse coefficients, a non-convex function is also introduced into the optimization problem. In addition, an iterative algorithm is developed to resolve the constructed sparse optimization problem, and the parameter selection method is also investigated to ensure the fault transient extraction ability of PCSR. The performance of the proposed PCSR is assessed via a synthetic and actual vibration signal. The results illustrate that the proposed PCSR has an excellent ability in fault transient extraction and machine fault detection.

Intelligent fault diagnosis for rotating machinery using L1/2-SF under variable rotational speed

Sparse filtering (SF), as an effective feature extraction technique, has attracted considerable attention in the field of mechanical fault diagnosis. But the generalization ability of SF to handle non-stationary signal under variable rotational speed is still poor. When the rotating parts of mechanical transmission work at a constant speed, the collected vibration signal is strongly correlated with the fault type. However, the mappings will no longer be so simple under the condition of variable rotational speed, which brings a rigorous challenge to intelligent fault diagnosis. To overcome the aforementioned deficiency, a novel L1/2 regularized SF method ( L1/2-SF) is studied in this paper. Specifically, L1/2 regularization strategy is added to the cost function of SF, then the L1/2-SF is directly employed to extract sparse features from the raw vibration data under variable rotational speed condition. In order to understand the sparse feature extraction ability of the L1/2 regularization, a physical explanation of the sparse solution generated by the L1/2 regularization strategy is explored. Next, softmax regression is employed for fault classification connected with the output layer of L1/2-SF. The effectiveness of L1/2-SF method is verified using a planetary gearbox dataset and a bearing dataset, respectively. Experiment results show that L1/2-SF can deal well with the variable rotational speed problem and is superior to other methods.

Combining adaptive sparse NMF feature extraction and soft mask to optimize DNN for speech enhancement

In masking-based deep neural network (DNN) speech enhancement, the time–frequency masking value cannot be estimated accurately because the potential structure information of speech is ignored. In this paper, a speech enhancement method is proposed by combining adaptive sparse non-negative matrix factorization (NMF) feature extraction and soft mask to optimize DNN, using the advantages of the sparse matrix in catching the protruding structure of speech and combining with optimized masking-based prediction. First, considering the dominance of speech and noise interference in different noisy speech signals, this paper proposes a new method for estimating soft mask value, and the initial soft mask value is estimated by using speech cochleagram and noise cochleagram. Then, speech cochleagram and noise cochleagram are learned separately by the sparse NMF (SNMF) to obtain a joint dictionary. The noisy speech is sparsely represented on the joint dictionary, and the adaptive adjustment factor related to the changes of speech and noise dictionary is added to obtain the sparse coefficient. The sparse coefficient is used as the input of the DNN model, and the initial soft mask value is used as the learning label to estimate the final soft mask value. Finally, the estimated soft mask value is combined with the noisy speech cochleagram to obtain enhanced speech. Compared with other methods, the results show that 1.6039 dB increases the average signal-to-noise ratio (SNR) of the proposed method, the average perceptual evaluation of speech quality (PESQ) is increased by 0.1994, and the average short-time objective intelligibility (STOI) is improved by 0.0271, which fully illustrate the superiority of the proposed algorithm.

A novel dictionary learning method based on total least squares approach with application in high dimensional biological data

In recent years dictionary learning has become a favorite sparse feature extraction technique. Dictionary learning represents each data as a sparse combination of atoms (columns) of the dictionary matrix. Usually, the input data is contaminated by errors that affect the quality of the obtained dictionary and so sparse features. This effect is especially critical in applications with high dimensional data such as gene expression data. Therefore, some robust dictionary learning methods have investigated. In this study, we proposed a novel robust dictionary learning algorithm, based on the total least squares, that could consider the inexactness of data in modeling. We confirm that standard and some robust dictionary learning models are the particular cases of our proposed model. Also, the results on various data indicate that our method performs better than other dictionary learning methods on high dimensional data.

Magnetic Anomaly Feature Extraction Using the Tunable Q-factor Wavelet Transform Based on Non-convex Overlapping Group Shrinkage

Extraction of the magnetic anomaly signal is one of the difficulties in the magnetic anomaly detection as the weak features extracted are easily disturbed by strong background noise. To address this problem, a sparse feature extraction method based on the tunable Q-factor wavelet transform and overlapping group shrinkage is developed in this paper. Compared with the traditional wavelet transform, the proposed method has excellent characteristics, which can flexibly tune the Q-factor according to the oscillation characteristics of the useful features. In this way, the sparsity of extracted features can be induced more effectively. In addition, the non-convex overlapping group shrinkage can effectively extract weak features from signals with group property, enhancing the extraction accuracy of features. The practical experiment verifies the effectiveness of the proposed method in the magnetic anomaly detection.

Quantized Constant-Q Gabor Atoms for Sparse Binary Representations of Cyber-Physical Signatures.

Increased data acquisition by uncalibrated, heterogeneous digital sensor systems such as smartphones present new challenges. Binary metrics are proposed for the quantification of cyber-physical signal characteristics and features, and a standardized constant-Q variation of the Gabor atom is developed for use with wavelet transforms. Two different continuous wavelet transform (CWT) reconstruction formulas are presented and tested under different signal to noise ratio (SNR) conditions. A sparse superposition of Nth order Gabor atoms worked well against a synthetic blast transient using the wavelet entropy and an entropy-like parametrization of the SNR as the CWT coefficient-weighting functions. The proposed methods should be well suited for sparse feature extraction and dictionary-based machine learning across multiple sensor modalities.

A novel sparse feature extraction method based on sparse signal via dual-channel self-adaptive TQWT

Sparse signal is a kind of sparsematrices which can carry fault information and simplify the signal at the same time. This can effectively reduce the cost of signal storage, improve the efficiency of data transmission, and ultimately save the cost of equipment fault diagnosis in the aviation field. At present, the existing sparse decomposition methods generally extract sparse fault characteristics signals based on orthogonal basis atoms, which limits the adaptability of sparse decomposition. In this paper, a self-adaptive atom is extracted by the improved dual-channel tunable Q-factor wavelet transform (TQWT) method to construct a self-adaptive complete dictionary. Finally, the sparse signal is obtained by the orthogonal matching pursuit (OMP) algorithm. The atoms obtained by this method are more flexible, and are no longer constrained to an orthogonal basis to reflect the oscillation characteristics of signals. Therefore, the sparse signal can better extract the fault characteristics. The simulation and experimental results show that the self-adaptive dictionary with the atom extracted from the dual-channel TQWT has a stronger decomposition freedom and signal matching ability than orthogonal basis dictionaries, such as discrete cosine transform (DCT), discrete Hartley transform (DHT) and discrete wavelet transform (DWT). In addition, the sparse signal extracted by the self-adaptive complete dictionary can reflect the time-domain characteristics of the vibration signals, and can more accurately extract the bearing fault feature frequency.

A SAR Image Target Recognition Approach via Novel SSF-Net Models.

With the wide application of high-resolution radar, the application of Radar Automatic Target Recognition (RATR) is increasingly focused on how to quickly and accurately distinguish high-resolution radar targets. Therefore, Synthetic Aperture Radar (SAR) image recognition technology has become one of the research hotspots in this field. Based on the characteristics of SAR images, a Sparse Data Feature Extraction module (SDFE) has been designed, and a new convolutional neural network SSF-Net has been further proposed based on the SDFE module. Meanwhile, in order to improve processing efficiency, the network adopts three methods to classify targets: three Fully Connected (FC) layers, one Fully Connected (FC) layer, and Global Average Pooling (GAP). Among them, the latter two methods have less parameters and computational cost, and they have better real-time performance. The methods were tested on public datasets SAR-SOC and SAR-EOC-1. The experimental results show that the SSF-Net has relatively better robustness and achieves the highest recognition accuracy of 99.55% and 99.50% on SAR-SOC and SAR-EOC-1, respectively, which is 1% higher than the comparison methods on SAR-EOC-1.

A Discrete-Time Neurodynamic Approach to Sparsity-Constrained Nonnegative Matrix Factorization.

Sparsity is a desirable property in many nonnegative matrix factorization (NMF) applications. Although some level of sparseness of NMF solutions can be achieved by using regularization, the resulting sparsity depends highly on the regularization parameter to be valued in an ad hoc way. In this letter we formulate sparse NMF as a mixed-integer optimization problem with sparsity as binary constraints. A discrete-time projection neural network is developed for solving the formulated problem. Sufficient conditions for its stability and convergence are analytically characterized by using Lyapunov's method. Experimental results on sparse feature extraction are discussed to substantiate the superiority of this approach to extracting highly sparse features.

Intelligent diagnosis of rolling bearing compound faults based on device state dictionary set sparse decomposition feature extraction–hidden Markov model

Identification of rolling bearing fault patterns, especially for the compound faults, has attracted notable attention and is still a challenge in fault diagnosis. Intelligent diagnosis method is an effective method for compound faults of rolling element bearing, and effective fault feature extraction is the key step to decide the intelligent diagnosis result to some extent. The sparse decomposition method could capture the complex impulsive characteristic components of rolling bearing more effectively than the other time–frequency analysis method when compound fault arises in rolling bearing. Based on the self-learning dictionary under different operating states of the device corresponding to the special features modes, an intelligent diagnosis method of rolling bearing compound faults based on device state dictionary set sparse decomposition feature extraction–hidden Markov model is proposed in the article. First, characteristic dictionaries of rolling bearing under different operating conditions are extracted by sparse decomposition self-learning method, and state dictionary set of rolling bearing is constructed. Then, the compound fault signals of bearing are transformed into sparse domain using the constructed dictionary set to extract sparse features. At last, the extracted sparse features are used as training and testing vectors of hidden Markov model, and satisfactory intelligent diagnosis results are obtained. The validity of the proposed method is verified by compound faults of rolling element bearing. In addition, the advantages of the proposed method are also verified by comparing with the other feature extraction and intelligent diagnosis methods, and the proposed method provides a feasible and efficient solution for fault diagnosis of rolling bearing compound faults.

Asymmetric penalty sparse model based cepstrum analysis for bearing fault detections

Signal processing approaches of rotating machine are well-established and wildly used in machine condition monitoring. Cepstrum analysis has the ability to transform related components from convolution form to addition form and suitable for the extraction of the periodic impulse components. However, the fault features are often localized in heavy background noises caused by the operation condition, which limit the application of cepstrum analysis. In this paper, a method is proposed by introducing the asymmetric penalty sparse model (APSM) to improve the effectiveness of cepstrum. The core idea of APSM is a sparse optimization formulation solved by the alternating direction method to achieve sparse feature extraction without any prior knowledge. The detection of bearing faults both in stationary and non-stationary conditions are given by using simulation and experimental signals. The performance of the proposed method is verified and the superiority is addressed by comparing with cesptrum analysis.

Structured optimal graph based sparse feature extraction for semi-supervised learning

Graph-based feature extraction is an efficient technique for data dimensionality reduction, and it has gained intensive attention in various fields such as image processing, pattern recognition, and machine learning. However, conventional graph-based dimensionality reduction algorithms usually depend on a fixed weight graph called similarity matrix, which seriously affects the subsequent feature extraction process. In this paper, a novel structured optimal graph based sparse feature extraction (SOGSFE) method for semi-supervised learning is proposed. In the proposed method, the local structure learning, sparse representation, and label propagation are simultaneously framed to perform data dimensionality reduction. In particular the similarity matrix and the projection matrix are obtained by an iterative calculation manner. The experimental results on several public image datasets demonstrate the robustness and effectiveness of the proposed method.

A SAR Target Recognition Based on Guided Reconstruction and Weighted Norm-Constrained Deep Belief Network

The deep learning has some new vitality to the synthetic aperture radar (SAR) automatic target recognition (ATR). By introducing different constraints, deep belief network (DBN) has become apply to SAR target recognition recent years, but the existing DBN algorithms have some questions including the high training epochs, low recognition rate and complex structure. Therefore, an algorithm based on guided reconstruction and weighted norm-constrained DBN is proposed. Firstly, in order to reduce the dimension of the image output feature, increase the speed of preprocessing, generate a one-dimensional image vector and normalized, a SAR target classification algorithm with two-scale fusion character based on guided filter reconstruction algorithm is introduced. Then, the sparse feature extraction of SAR image is carried out by weighted norm-constrained DBN. By the regularization constraint of the probability distribution, the algorithm can minimize the joint probability distribution of visual layer and hidden layer by samples. The low-dimensional feature is further improved based on the generalized optimization of norm-constrained RBM. Finally, by a regularized Softmax which can classify the targets and obtain output results. The experimental results show that the SAR target recognition algorithm based on the guided reconstruction and weighted constrained deep confidence network not only improves the target recognition performance and generalization ability, but also reduces the output feature dimension and network training times, and the recognition performance of the algorithm is further improved.

A deep learning algorithm using a fully connected sparse autoencoder neural network for landslide susceptibility prediction

The environmental factors of landslide susceptibility are generally uncorrelated or non-linearly correlated, resulting in the limited prediction performances of conventional machine learning methods for landslide susceptibility prediction (LSP). Deep learning methods can exploit low-level features and high-level representations of information from environmental factors. In this paper, a novel deep learning–based algorithm, the fully connected spare autoencoder (FC-SAE), is proposed for LSP. The FC-SAE consists of four steps: raw feature dropout in input layers, a sparse feature encoder in hidden layers, sparse feature extraction in output layers, and classification and prediction. The Sinan County of Guizhou Province in China, with a total of 23,195 landslide grid cells (306 recorded landslides) and 23,195 randomly selected non-landslide grid cells, was used as study case. The frequency ratio values of 27 environmental factors were taken as the input variables of FC-SAE. All 46,390 landslide and non-landslide grid cells were randomly divided into a training dataset (70%) and a test dataset (30%). By analyzing real landslide/non-landslide data, the performances of the FC-SAE and two other conventional machine learning methods, support vector machine (SVM) and back-propagation neural network (BPNN), were compared. The results show that the prediction rate and total accuracies of the FC-SAE are 0.854 and 85.2% which are higher than those of the SVM-only (0.827 and 81.56%) and BPNN (0.819 and 80.86%), respectively. In conclusion, the asymmetric and unsupervised FC-SAE can extract optimal non-linear features from environmental factors successfully, outperforms some conventional machine learning methods, and is promising for LSP.

Sparse coding-based feature extraction for biometric remote authentication in Internet of Things

In recent years, Internet of Things (IoT) has attracted lots of attention. However, the security related issues such as authentication remain a challenge. The heterogeneity of IoT in terms of devices and communication makes most existing authentication mechanisms inapplicable. So, there is a need for a two-factor authentication mechanism to obtain an end-to-end authentication between IoT devices/applications. In this paper, we propose a sparse coding based feature extraction for biometric remote user authentication. The proposed scheme makes use of sparse codes hash operations and overcomplete dictionary to store/retrieve the biometric data efficiently. The performance analysis proves that the proposed method is robust against noise and able to obtain the accuracy of 0.97.

Multi-objective data-ensemble wind speed forecasting model with stacked sparse autoencoder and adaptive decomposition-based error correction

Accurate wind speed prediction is essential for proper use of wind energy resource. In this paper, a novel hybrid multi-step wind speed forecasting model is developed, which consists of sparse feature extraction, bidirectional deep learning, multi-objective optimization, and adaptive decomposition-based error correction. Apart from the traditional average-based resolution transformation method, a two-layer stacked sparse autoencoder (SSAE) is proposed to extract the hidden representation of original 3 s high-resolution wind speed data. Trained by the data generated from different resolution transformation methods, two bidirectional long short-term memory (BiLSTM) networks serve as base predictors and provide 10-step forecasting results. The results of base predictors are reasonably ensembled by multi-objective multi-universe optimization (MOMVO). Moreover, to reduce the predictable components in error series further, a correction model based on empirical wavelet transform (EWT) and outlier robust extreme learning machine model (ORELM) is constructed to reduce the forecasting error further. The effectiveness of the proposed hybrid model is comprehensively evaluated by a series of experiments. The experimental results demonstrate that: (a) the proposed model is well trained, with great convergence, and an average RMSE of 0.2618 m/s in 10-step forecasting; (b) the proposed model outperforms other existing models in all experimental sites and forecasting steps; (c) the multi-objective optimization algorithm can rationally integrate base predictors to obtain better performance in each step; (d) the proposed residual error correction model can generate more than 78% improvement of RMSE, significantly better than compared correction methods.

Reduced iteration image reconstruction of incomplete projection CT using regularization strategy through Lp norm dictionary learning.

For sparse and limited angle projection Computed Tomography (CT), the reconstructed image usually suffers from considerable artifacts due to undersampled data. To improve image reconstruction quality of sparse and limited angle projection CT, this study tested a novel reconstruction algorithm based on Dictionary Learning (DL) from sparse and limited projections. The study used signal sparse representation and feature extraction to render the DL technology, which is constrained by L2 and Lp norms, respectively. A Lp Norm Dictionary Learning term is suitable for regular term of objective function for CT image reconstruction. This is helpful for solving the objective function by combining algorithm of ART. Based on these features, the new algorithm of ART-DL-Lp is proposed for CT image reconstruction. The alternate solving strategy of the algorithm of "ART first, then adaptive DL" is provided in sequence. The impact on reconstruction results of ART-DL-Lp at different p values (0 < p < 1) is also considered. For non-ideal projections with noise, the digital experiments show that ART-DL-Lp data were superior to those of ART, SART, and ART-DL-L2. Accordingly, the objective evaluation metrics for non-ideal situation of RMSE, MAE, PSNR, Residuals and SSIM are all better than those of contrasted three algorithms. The metrics curves of ART-DL-Lp algorithm are recorded as the best. In both incomplete projection situations, smaller p-value of ART-DL-Lp algorithm induces more close reconstructed images to the original form and better five objective evaluation metrics. Overall, the reconstruction efficiency of the proposed ART-DL-Lp for CT imaging using the noisy incomplete projections outperforms ART, SART and ART-DL-L2 algorithms. For ART-DL-Lp algorithm, lower p-values result in better reconstruction performance.