Fast Learning Method Research Articles

Aplikasi Metode al Miftah Dalam Mendukung Pembelajaran Bahasa Arab

Arabic is a language that must be learned by all Muslims in the world. Studying Arabic as a second language will certainly encounter significant difficulties. The differences in terms in Arabic mean that most students do not understand how to use them. Fast learning methods in learning Arabic are increasingly developing. One of them is the al-miftah method. The al-miftah method is used to equate terms in Arabic to Indonesian so that they are easy to understand. For this reason, in this research the researcher wants to combine the al-miftah method into the Arabic language learning process in class VII of Mts KHA Wahid Hasyim Bangil

Deep learning for fast photoacoustic wave simulations

Photoacoustic tomography involves absorption of pulsed light and subsequent generation of ultrasound, which when detected using an array of sensors can produce clinically useful images. Simulation tools for photoacoustic wave propagation have played a key role in advancing photoacoustic imaging by providing quantitative and qualitative insights into parameters affecting image quality. Classical methods for numerically solving the photoacoustic wave equation rely on a fine discretization of space and can become computationally expensive for large computational grids. In this work, we apply Fourier Neural Operator (FNO) networks as a fast data-driven deep learning method for solving the 2D photoacoustic wave equation in a homogeneous medium. Comparisons between the FNO network and pseudo-spectral time domain approach demonstrated that the FNO network generated comparable simulations with small errors and was an order of magnitude faster. Moreover, the FNO network was generalizable and could generate simulations not observed in the training data.

TPE-RBF-SVM Model for Soybean Categories Recognition in Selected Hyperspectral Bands Based on Extreme Gradient Boosting Feature Importance Values

Soybeans with insignificant differences in appearance have large differences in their internal physical and chemical components; therefore, follow-up storage, transportation and processing require targeted differential treatment. A fast and effective machine learning method based on hyperspectral data of soybeans for pattern recognition of categories is designed as a non-destructive testing method in this paper. A hyperspectral-image dataset with 2299 soybean seeds in four categories is collected. Ten features are selected using an extreme gradient boosting algorithm from 203 hyperspectral bands in a range of 400 to 1000 nm; a Gaussian radial basis kernel function support vector machine with optimization by the tree-structured Parzen estimator algorithm is built as the TPE-RBF-SVM model for pattern recognition of soybean categories. The metrics of TPE-RBF-SVM are significantly improved compared with other machine learning algorithms. The accuracy is 0.9165 in the independent test dataset, which is 9.786% higher for the vanilla RBF-SVM model and 10.02% higher than the extreme gradient boosting model.

A fast sparse Bayesian learning method with adaptive Laplace prior for space‐time adaptive processing

Space-time adaptive processing with finite samples is supposed to be a crucial technique for airborne radar systems. Inspired by the application of Gaussian prior in sparse Bayesian learning algorithm and the adaptive least absolute shrinkage and selection operator algorithm, a hierarchical Bayesian framework with adaptive Laplace priors is proposed. In this paper, a novel method is applied to avoid the high-dimension matrix inverse operation in the proposed algorithm. Moreover, in order to apply the method in the complex-valued domain, the complex-valued signal is split into two independent variables. Then, the sparse recovery problem in the complex-valued domain can be transformed into the real-value domain. Simulation experiments show that the proposed algorithm can achieve great clutter suppression performance and also ensure high computational efficiency.

Simple framework for the contrastive learning of visual representations-based data-driven tight frame for seismic denoising and interpolation

Complex seismic data sets benefit from adaptive sparse transformation methods to improve resolution and fidelity. The data-driven tight frame (DDTF) is a fast dictionary learning method that can simplify the sparse representation process and efficiently update the compact frame. However, the DDTF-based method is computationally intensive and relatively ineffective in accurately characterizing high-dimensional data sets with weak features and high noise levels. A new adaptive patch-selection method based on a simple framework for the contrastive learning of visual representations (SimCLR) is designed to further accelerate and improve the dictionary training process in the DDTF. The variance of all of the available patches is first calculated; then, the cosine similarity between the patch with the maximum variance and the other patches is calculated. Finally, the patches with large similarities are used for the filter bank training process. This method can efficiently track seismic events. The efficacy of applying the trained filter bank for seismic data recovery is tested by using the Monte Carlo-DDTF and SimCLR-DDTF methods. The numerical results demonstrate that the SimCLR-DDTF method can intelligently select training patches with more effective information, greatly improve the training efficiency of the DDTF, and obtain high-quality denoising and interpolation results.

A multi-birth metric learning framework based on binary constraints

Multi-metric learning plays a significant role in improving the generalization of algorithms related to distance metrics since using a single metric is sometimes insufficient to handle complex data. Metric learning can adjust automatically the distance between samples to make the intra-class samples compact while making the inter-class distance as far as possible. To implement this intention better,in this work, we propose a novel multi-metric learning framework based on the pair constraints instead of triple constraints to reduce computational burden. To solve effectively the problem, we first propose a multi-birth metric learning model (termed MBML), where for each class sample, the global metric and a local metric are jointly trained. Both global and local structural information are adapted to better depict sample information. Then two alternating iterative algorithms are developed to optimize the MBML. The convergence of the proposed algorithm and complexity are analyzed theoretically. Moreover, a fast diagonal multi-metric learning method is proposed based on binary constraints, and problem can be reformulated a linear programming, with fast training speed, low the computational burden and the global optimal solutions. Numerical experiments are carried out on different scales and different types of datasets including an artificial data, benchmark datasets and an image database from binary class and multi-class problems. Experiment results confirm the feasibility and effectiveness of the proposed methods.

Initial Stage COVID-19 Detection System Based on Patients’ Symptoms and Chest X-Ray Images

ABSTRACT The accurate diagnosis of the initial stage COVID-19 is necessary for minimizing its spreading rate. The physicians most often recommend RT-PCR tests; this is invasive, time-consuming, and ineffective in reducing the spread rate of COVID-19. However, this can be minimized by using noninvasive and fast machine learning methods trained either on labeled patients’ symptoms or medical images. The machine learning methods trained on labeled patients’ symptoms cannot differentiate between different types of pneumonias like COVID-19, viral pneumonia, and bacterial pneumonia because of similar symptoms, i.e., cough, fever, headache, sore throat, and shortness of breath. The machine learning methods trained on labeled patients’ medical images have the potential to overcome the limitation of the symptom-based method; however, these methods are incapable of detecting COVID-19 in the initial stage because the infection of COVID-19 takes 3 to 12 days to appear. This research proposes a COVID-19 detection system with the potential to detect COVID-19 in the initial stage by employing deep learning models over patients’ symptoms and chest X-Ray images. The proposed system obtained average accuracy 78.88%, specificity 94%, and sensitivity 77% on a testing dataset containing 800 patients’ X-Ray images and 800 patients’ symptoms, better than existing COVID-19 detection methods.

Fast Burst-Sparsity Learning-Based Baseline Correction (FBSL-BC) Algorithm for Signals of Analytical Instruments.

Baseline correction is a critical step for eliminating the interference of baseline drift in spectroscopic analysis. The recently proposed sparse Bayesian learning (SBL)-based method can significantly improve the baseline correction performance. However, it has at least two disadvantages: (i) it works poorly for large-scale datasets and (ii) it completely ignores the burst-sparsity structure of the sparse representation of the pure spectrum. In this paper, we present a new fast burst-sparsity learning method for baseline correction to overcome these shortcomings. The first novelty of the proposed method is to jointly adopt a down-sampling strategy and construct a multiple measurements block-sparse recovery problem with the down-sampling sequences. The down-sampling strategy can significantly reduce the dimension of the spectrum; while jointly exploiting the block sparsity among the down-sampling sequences avoids losing the information contained in the original spectrum. The second novelty of the proposed method is introducing the pattern-coupled prior into the SBL framework to characterize the inherent burst-sparsity in the sparse representation of spectrum. As illustrated in the paper, burst-sparsity commonly occurs in peak zones with more denser nonzero coefficients. Properly utilizing such burst-sparsity can further enhance the baseline correction performance. Results on both simulated and real datasets (such as FT-IR, Raman spectrum, and chromatography) verify the substantial improvement, in terms of estimation accuracy and computational complexity.

Large-Scale Multi-View Clustering via Fast Essential Subspace Representation Learning

Large-scale Multi-View Clustering (LMVC) is a hot research problem in the fields of signal processing and machine learning, and many anchor-based multi-view subspace clustering algorithms are proposed in recent years. However, most existing methods usually concentrate on the issue of reducing the time cost and ignore the exploration of the complementary information during the clustering process. To this end, we propose a Fast Essential Subspace Representation Learning (FESRL) method for large-scale multi-view subspace clustering. Specifically, FESRL introduces the orthogonal transformation to investigate both the complementary and consensus information across multiple views. The essential subspace representation can be learned in a linear time cost. Experiments conducted on several benchmark datasets illustrate the competitiveness of the proposed method.

Fast dictionary learning for 3D simultaneous seismic data reconstruction and denoising

Simultaneous seismic data reconstruction and denoising is a hot research topic. The sparse representation method based on dictionary learning is one of the most effective methods to reconstruct seismic data and suppress noise. Sparse dictionary traditionally uses the K-means singular value decompositions (K-SVD) method for learning. However, the main disadvantage of K-SVD is that it requires many singular value decompositions (SVDs), which is low in computational efficiency and not suitable for practical applications, especially in high-dimensional problems. To address the computational efficiency problem of K-SVD, we propose a fast dictionary learning method based on sequence generalized K-means (SGK) algorithm for efficient reconstruction and denoising of 3D seismic data. In SGK algorithm, dictionary atoms are updated by arithmetic average of several training signals instead of singular value decomposition in the K-SVD algorithm. The performance of the two methods is verified by 3D numerical examples. The results demonstrate that the proposed reconstruction and denoising method using SGK can achieve comparable performance as the K-SVD method but significantly improve the computational efficiency.

LISA2: Learning Complex Single-Cell Trajectory and Expression Trends.

Single-cell transcriptional and epigenomics profiles have been applied in a variety of tissues and diseases for discovering new cell types, differentiation trajectories, and gene regulatory networks. Many methods such as Monocle 2/3, URD, and STREAM have been developed for tree-based trajectory building. Here, we propose a fast and flexible trajectory learning method, LISA2, for single-cell data analysis. This new method has two distinctive features: (1) LISA2 utilizes specified leaves and root to reduce the complexity for building the developmental trajectory, especially for some special cases such as rare cell populations and adjacent terminal cell states; and (2) LISA2 is applicable for both transcriptomics and epigenomics data. LISA2 visualizes complex trajectories using 3D Landmark ISOmetric feature MAPping (L-ISOMAP). We apply LISA2 to simulation and real datasets in cerebellum, diencephalon, and hematopoietic stem cells including both single-cell transcriptomics data and single-cell assay for transposase-accessible chromatin data. LISA2 is efficient in estimating single-cell trajectory and expression trends for different kinds of molecular state of cells.

Fast local representation learning via adaptive anchor graph for image retrieval

Linear Discriminant Analysis (LDA) is one of most important methods in dimensionality reduction domain, which is limited with Gaussian assumption. Because of the complexity of real data, data often presents non-Gaussian distribution that points in same class can be divided into several sub-clusters and center point is not enough to describe the distribution of data. In order to solve non-Gaussian data, LDA-based methods consider local structure information through measuring each pairwise distance of full connection graph. However, the strategy of establishing fully-connected graph is at expense of high computational complexity and limits it practical and industrial applications. We propose a Fast Local Representation Learning (FLRL) method which leverages anchor points to establish anchor-based graph and uses similarity matrix to depict the relationships of each pairwise connections. Notably, to avoid the affect of noises and redundant features in original space, anchor points and similarity matrix are updated alternately in subspace that local structure of data will be more precise to learn. Extensive pattern classification and image retrieval experiments on several synthetic datasets, well-known datasets and deep features datasets demonstrate the advantages of our method over the state-of-the-art methods. Our source code available on:https://github.com/superzcy/FLRL.

Transfer-RLS method and transfer-FORCE learning for simple and fast training of reservoir computing models

Reservoir computing is a machine learning framework derived from a special type of recurrent neural network. Following recent advances in physical reservoir computing, some reservoir computing devices are thought to be promising as energy-efficient machine learning hardware for real-time information processing. To realize efficient online learning with low-power reservoir computing devices, it is beneficial to develop fast convergence learning methods with simpler operations. This study proposes a training method located in the middle between the recursive least squares (RLS) method and the least mean squares (LMS) method, which are standard online learning methods for reservoir computing models. The RLS method converges fast but requires updates of a huge matrix called a gain matrix, whereas the LMS method does not use a gain matrix but converges very slow. On the other hand, the proposed method called a transfer-RLS method does not require updates of the gain matrix in the main-training phase by updating that in advance (i.e., in a pre-training phase). As a result, the transfer-RLS method can work with simpler operations than the original RLS method without sacrificing much convergence speed. We numerically and analytically show that the transfer-RLS method converges much faster than the LMS method. Furthermore, we show that a modified version of the transfer-RLS method (called transfer-FORCE learning) can be applied to the first-order reduced and controlled error (FORCE) learning for a reservoir computing model with a closed-loop, which is challenging to train.

A fast active learning method in design of experiments: multipeak parallel adaptive infilling strategy based on expected improvement

A fast active learning method in design of experiments: multipeak parallel adaptive infilling strategy based on expected improvement

An Accurate Tool for Uncovering Cancer Subtypes by Fast Kernel Learning Method to Integrate Multiple Profile Data.

In recent years, cancer has become a severe threat to human health. If we can accurately identify the subtypes of cancer, it will be of great significance to the research of anti-cancer drugs, the development of personalized treatment methods, and finally conquer cancer. In this paper, we obtain three feature representation datasets (gene expression profile, isoform expression and DNA methylation data) on lung cancer and renal cancer from the Broad GDAC, which collects the standardized data extracted from The Cancer Genome Atlas (TCGA). Since the feature dimension is too large, Principal Component Analysis (PCA) is used to reduce the feature vector, thus eliminating the redundant features and speeding up the operation speed of the classification model. By multiple kernel learning (MKL), we use Kernel target alignment (KTA), fast kernel learning (FKL), Hilbert-Schmidt Independence Criterion (HSIC), Mean to calculate the weight of kernel fusion. Finally, we put the combined kernel function into the support vector machine (SVM) and get excellent results. Among them, in the classification of renal cell carcinoma subtypes, the maximum accuracy can reach 0.978 by using the method of MKL (HSIC calculation weight), while in the classification of lung cancer subtypes, the accuracy can even reach 0.990 with the same method (FKL calculation weight).

A Blood Glucose Control Framework Based on Reinforcement Learning With Safety and Interpretability: In Silico Validation

Controlling blood glucose levels in diabetic patients is important for managing their health and quality of life. Several algorithms based on model predictive control and reinforcement learning (RL) have been proposed so far, most of which use prior knowledge of physiological systems, the mathematical structure of blood glucose dynamics, and many episodes including failures for training the policy network in RL. To be smoothly adopted in clinical settings, we propose a fast online learning method underlining safety and interpretability. A random forest regressor and a dual attention network were exploited for glucose prediction and extension of state variables. The soft actor-critic network to determine insulin dosing was guided by proportional-integral-derivative (PID) control in the early phase, and an adaptive safe actor with suspension and additional insulin dosing was incorporated. The performance of the models was validated using an FDA-approved type 1 diabetes simulator. The results showed comparable outcomes with PID control. Using this system, glucose dynamics could be captured despite minimal prior knowledge. The extended state variables were correlated with basic states such as glucose, insulin, and meal intake, their derivatives, and their integrals, which can be fundamental elements of mathematical modeling of physiological responses. Attention scores and attribution scores in the prediction and control models represented the focused features and the internal operation of the models with interpretability. We expect this study to provide some insights on how RL can be practically adopted in clinical environments and how interpretability can provide hints of machines' thoughts for clinical applications.

A fast and accurate deep learning method for strawberry instance segmentation

Fruit instance segmentation is a key element for autonomous picking systems. This paper proposes a methodology for instance segmentation of strawberries using deep learning techniques. This methodology presents the following changes in relation to the well-known Mask R-CNN network: a new architecture is designed for the backbone and the mask network; the object classifier and the bounding-box regressor are removed; the non-maximum suppression algorithm is replaced with a new region grouping and filtering algorithm without increasing the order of complexity.For this research, a new high-resolution data set of 3100 strawberry images is obtained, along with the corresponding manually-segmented ground-truth images (StrawDI_Db1 database). In addition to using widely-used metrics based on Average Precision (AP), a new performance metric is proposed: the Instance Intersection Over Union (I2oU), to assess different options of instance segmentation techniques. This material, which is published and available to the research community, allows the rigorous performance comparison between future methodologies.The results obtained in the 200 images included in the test set show that the proposed methodology significantly reduces the inference time (10 fps vs. 5 fps) while maintaining competitive results to the original Mask R-CNN for mean AP (43.85 vs. 45.36) and mean I2oU (87.27 vs. 87.70) metrics. Thus, the proposed method can be considered as the most promising one in view of its possible integration in automatic strawberry picking systems.

PREASC

The increasing functionality of electronic systems due to the constant evolution of the market requirements makes the non-functional aspects of such systems (e.g., energy consumption, area overhead, or performance) a major concern in the design process. Approximate computing is a promising way to optimize these criteria by trading accuracy within acceptable limits. Since the cost of applying significant structural changes to a given design increases with the stage of development, the optimization solution needs to be incorporated into the design as early as possible. For the early design entry, modeling hardware at the Electronic System Level (ESL) using the SystemC language is nowadays widely used in the industry. To apply approximation techniques to optimize a given SystemC design, designers need to know which parts of the design can be approximated. However, identifying these parts is a crucial and non-trivial starting point of approximate computing, as the incorrect detection of even one critical part as resilient may result in an unacceptable output. This usually requires a significant programming effort by designers, especially when exploring the design space manually. In this article, we present PREASC, a fully automated framework to identify the resilience portions of a given SystemC design. PREASC is based on a combination of static and dynamic analysis methods along with regression analysis techniques (a fast machine learning method providing an accurate function estimation). Once the resilient portions are identified, an approximation degree analysis is performed to determine the maximum error rate that each resilient portion can tolerate. Subsequently, the maximum number of resilient portions that can be approximated at the same time are reported to designers at different granularity levels. The effectiveness of our approach is evaluated using several standard SystemC benchmarks from various domains.

An Analysis of Fast Learning Methods for Classifying Forest Cover Types

ABSTRACT Proper mapping and classification of Forest cover types are integral in understanding the processes governing the interaction mechanism of the surface with the atmosphere. In the presence of massive satellite and aerial measurements, a proper manual categorization has become a tedious job. In this study, we implement three different modest machine learning classifiers along with three statistical feature selectors to classify different cover types from cartographic variables. Our results showed that, among the chosen classifiers, the standard Random Forest Classifier together with Principal Components performs exceptionally well, not only in overall assessment but across all seven categories. Our results are found to be significantly better than existing studies involving more complex Deep Learning models.

A Novel Multi-Feature Joint Learning Method for Fast Polarimetric SAR Terrain Classification

Polarimetric synthetic aperture radar (PolSAR) image classification is one of the most important study areas for PolSAR image processing. Many kinds of PolSAR features can be extracted for PolSAR image classification, such as the scattering, polarimetric or image features. However, it is difficult to improve the classification accuracy of PolSAR images by using all these low-level features directly, since they may conflict with each other for classification. Hence, how to joint learn these low-level features to obtain high-level discriminating features is a challenging task. To solve this problem, a novel fast multi-feature joint learning method(fMF-JLC) is proposed for PolSAR image classification. The proposed method extract three kinds of low-level features of PolSAR data at first. Then, a multi-feature joint sparse representation model(MF-JSR) is proposed by designing joint sparse constraints on the extracted features above. Moreover, the joint sparse features are further compressed to overcome the dimension curse and acquire semantic features by the topic model. By this way, the low-level features are fused and discriminating high-level features are acquired. However, the pixel-wise feature learning method is time consuming. To speed the proposed method, a superpixel-based fast learning method is designed by involving the contextual relationship. Experiments are taken on three sets of real PolSAR data with different sensors and bands, and several compared methods are used to verify the effectiveness of the proposed method. The experimental results illustrate that the proposed method can obtain better performance than the state-of-art methods, especially for the heterogenous areas.