Search

Objectives. To develop an imputation method to produce estimates for suppressed values within a shared government administrative data set to facilitate accurate data sharing and statistical and spatial analyses. Methods. We developed an imputation approach that incorporated known features of suppressed Massachusetts surveillance data from 2011 to 2017 to predict missing values more precisely. Our methods for 35 de-identified opioid prescription data sets combined modified previous or next substitution followed by mean imputation and a count adjustment to estimate suppressed values before sharing. We modeled 4 methods and compared the results to baseline mean imputation. Results. We assessed performance by comparing root mean squared error (RMSE), mean absolute error (MAE), and proportional variance between imputed and suppressed values. Our method outperformed mean imputation; we retained 46% of the suppressed value’s proportional variance with better precision (22% lower RMSE and 26% lower MAE) than simple mean imputation. Conclusions. Our easy-to-implement imputation technique largely overcomes the adverse effects of low count value suppression with superior results to simple mean imputation. This novel method is generalizable to researchers sharing protected public health surveillance data. (Am J Public Health. 2021; 111(10):1830–1838. https://doi.org/10.2105/AJPH.2021.306432 )

Nonoccurring behavior (NOB) studies have attracted the growing attention of scholars as a crucial part of behavioral science. As an effective method to discover both NOB and occurring behaviors (OB), negative sequential pattern (NSP) mining is successfully used in analyzing medical treatment and abnormal behavior patterns. At this time, NSP mining is still an active and challenging research domain. Most of the algorithms are inefficient in practice. Briefly, the key weaknesses of NSP mining are: 1) an inefficient positive sequential pattern (PSP) mining process, 2) a strict constraint of negative containment, and 3) the lack of an effective Negative Sequential Candidate (NSC) generation method. To address these weaknesses, we propose a highly efficient algorithm with improved techniques, named sc-NSP, to mine NSP efficiently. We first propose an improved PrefixSpan algorithm in the PSP mining process, which connects to a bitmap storage structure instead of the original structure. Second, sc-NSP loosens the frequency constraint and exploits the NSC generation method of positive and negative sequential patterns mining (PNSP) (a classic NSP mining method). Furthermore, a novel pruning strategy is designed to reduce the computational complexity of sc-NSP. Finally, sc-NSP obtains the support of NSC by using the most efficient bitwise-based calculation operation. Theoretical analyses show that sc-NSP performs particularly well on data sets with a large number of elements and items in sequence. Comparison and extensive experiments along with case studies on health data show that sc-NSP is 10 times more efficient than other state-of-the-art methods, and the number of NSPs obtained is 5 times greater than other methods.

Gesture Recognition which is an inevitable part of human computer interaction is an ever evolving active research area. This has been in use under varied applications like smart home systems, sign language recognition, augment reality and in device controls. The acquisition of hand gestures are usually through optical sensors, in this work a radar based hand gesture data set is used for classifying the gestures. Micro doppler signatures were used as the input to the model. Radar dataset has fewer data samples compared to optical based data sets. The proposed work uses separable convolutional neural networks model which does depth wise convolution followed by point wise convolution to reduce overfitting effect of the training data. The proposed model was built in such a way that the model is capable of classifying any unseen data without exactly mimicking the training samples. The proposed model has achieved 94.56% as the testing accuracy which is certainly better than the previous work on this Dop Net data set. Moreover the model has also minimized the computational hours of the model using separable convolutions.

A new stochastic coordinate descent deep learning architectures optimization is proposed for Automated Diabetic Retinopathy Detection and Classification from different data sets and convolution networks. Initially, the layer-by-layer comparison of convolution matrix, pooling, transition, and dense of each network along with their matrix order is examined. Loss minimization after the testing data from the prediction analysis is considered as objective function for every stage of convolution networks. Similarity among the networks is identified for optimizing the matrix order of each layer and the minimization is performed for classifying the Diabetic Retinopathy level. The performance of the proposed optimization architectures confirmed through confusion matrix for every data set taking training accuracy and data loss as measures. The result of the proposed optimized schemes shows that for different the datasets different in amount data displays the dominance than existing schemes. The numerical analysis of different training accuracy and data loss is presented for validating the proposed work.

This paper will go through an attempt to predict vital data namely BPM and pulse interval of human being. Study involves data set produced with a total of 83 subjects spread across various categories by age group and BMI ratio. A specialized device fabricated for this purpose designed on a development board divided in two layers of printed circuit board based on Atmega328p microcontroller enables the collection of data using a pulse sensor. The machine learning algorithm running on the device saves encoded data files and produces calculation based on the previous data saved. Thus, the accuracy improves after every data sample obtained enabling the device with experience-based learning. The system makes use of a few mathematical equations on the dataset obtained for each category and is successfully able to predict BPM and pulse interval more than 70% times for age group 21 to 25 involving proportionate subjects according to BMI ratio with a database of 140 samples for each parameter taken from 70 subjects; predict BPM more than 50% times and pulse interval more than 30% times for age group 51 to 55 also involving proportionate subjects according to BMI ratio with a database of 26 samples for each parameter taken from 13 subjects while subjects were sitting or standing idle represented by the class “Regular”.

In today's competitive world, it is critical for an institute to forecast student performance, classify individuals based on their talents, and attempt to enhance their performance in future tests. Students should be advised well in advance to concentrate their efforts in a specific area in order to improve their academic achievement. This type of analysis assists an institute in lowering its failure rates. Based on their prior performance in comparable courses, this study predicts students' performance in a course. Data mining is a collection of techniques used to uncover hidden patterns in massive amounts of existing data. These patterns may be valuable for analysis and prediction. Education data mining refers to the collection of data mining applications in the field of education. These applications are concerned with the analysis of data from students and teachers. The analysis might be used for categorization or prediction. Machine learning such as Nave Bayes, ID3, C4.5, and SVM are investigated. UCI machinery student performance data set is used in experimental study. Algorithms are analysed on certain parameters like- accuracy, error rate.

Cotton detection is the localization and identification of the cotton in an image. It has a wide application in robot harvesting. Various modern algorithms use deep learning techniques for detection of fruits/flowers. As per the survey, the topics travelled include numerous algorithms used, and accuracy obtained on using those algorithms on their data set. The limitations and the advantages in each paper, are also discussed. This paper focuses on various fruit detection algorithms- the Faster RCNN, the RCNN, YOLO. Ultimately, a rigorous survey of many papers related to the detection of objects like fruits/flowers, analysis of the assets and faintness of each paper leads us to understanding the techniques and purpose of algorithms.

To identify the abnormal characteristics of the oplegnathus punctatus is great importance to the detection of iridovirus disease in the breeding environment. In this paper, an advanced neural network model to identify the characteristics of the oplegnathus punctatus and predict its different periods of suffering from iridovirus disease is proposed based on the establishment of a data set. First of all, a standard format data set of oplegnathus punctatus and an abnormal format date set are established in order to verify the effectiveness of the method in this paper. And then, the feature extraction fusion method is used for preprocessing in terms of the abnormal format data set, which combines the edge features extracted by the improved multi-template Sobel operator and the color features extracted by the HSV model. Finally, an improved VGG-GoogleNet network recognition model comes into being through the fusion and improvement of the VGG and GoogleNet neural network structure. The experiments results show that the prediction accuracy rate for oplegnathus punctatus suffering from iridovirus disease in the the abnormal format data set and the standard format data set are improved, which reach 98.55% and 69.18%.

The SLAM problem in dynamic scenes is regarded as a challenge. This article proposes a novel SLAM framework for dynamic environments, which combines neural network and motion information of dynamic objects, making the system more adaptable to dynamic scenes. Specifically, we adopt a fast object detection network, tightly couple the results of the object detection with the geometric information in the SLAM system. Then the feature point extracted from the image is associated with a dynamic probability. By utilizing the feature points which tend to be static, the localization result can be greatly improved in dynamic environments. We perform the experiments both on the public data set and the real environment. The result can demonstrate that the proposed method greatly improves the localization accuracy in a dynamic environment. An open-source version of the source code is available.

It is of great significance to apply deep learning for the early diagnosis of Alzheimer's disease (AD). In this work, a novel tensorizing GAN with high-order pooling is proposed to assess mild cognitive impairment (MCI) and AD. By tensorizing a three-player cooperative game-based framework, the proposed model can benefit from the structural information of the brain. By incorporating the high-order pooling scheme into the classifier, the proposed model can make full use of the second-order statistics of holistic magnetic resonance imaging (MRI). To the best of our knowledge, the proposed Tensor-train, High-order pooling and Semisupervised learning-based GAN (THS-GAN) is the first work to deal with classification on MR images for AD diagnosis. Extensive experimental results on Alzheimer's disease neuroimaging initiative (ADNI) data set are reported to demonstrate that the proposed THS-GAN achieves superior performance compared with existing methods, and to show that both tensor-train and high-order pooling can enhance classification performance. The visualization of generated samples also shows that the proposed model can generate plausible samples for semisupervised learning purpose.

Benefit from avoiding the utilization of labeled samples, which are usually insufficient in the real world, unsupervised learning has been regarded as a speedy and powerful strategy on clustering tasks. However, clustering directly from primal data sets leads to high computational cost, which limits its application on large-scale and high-dimensional problems. Recently, anchor-based theories are proposed to partly mitigate this problem and field naturally sparse affinity matrix, while it is still a challenge to get excellent performance along with high efficiency. To dispose of this issue, we first presented a fast semisupervised framework (FSSF) combined with a balanced K -means-based hierarchical K -means (BKHK) method and the bipartite graph theory. Thereafter, we proposed a fast self-supervised clustering method involved in this crucial semisupervised framework, in which all labels are inferred from a constructed bipartite graph with exactly k connected components. The proposed method remarkably accelerates the general semisupervised learning through the anchor and consists of four significant parts: 1) obtaining the anchor set as interim through BKHK algorithm; 2) constructing the bipartite graph; 3) solving the self-supervised problem to construct a typical probability model with FSSF; and 4) selecting the most representative points regarding anchors from BKHK as an interim and conducting label propagation. The experimental results on toy examples and benchmark data sets have demonstrated that the proposed method outperforms other approaches.

Recently, deep learning-based approaches have achieved superior performance on object detection applications. However, object detection for industrial scenarios, where the objects may also have some structures and the structured patterns are normally presented in a hierarchical way, is not well investigated yet. In this work, we propose a novel deep learning-based method, hierarchical graphical reasoning (HGR), which utilizes the hierarchical structures of trains for train component detection. HGR contains multiple graphical reasoning branches, each of which is utilized to conduct graphical reasoning for one cluster of train components based on their sizes. In each branch, the visual appearances and structures of train components are considered jointly with our proposed novel densely connected dual-gated recurrent units (Dense-DGRUs). To the best of our knowledge, HGR is the first kind of framework that explores hierarchical structures among objects for object detection. We have collected a data set of 1130 images captured from moving trains, in which 17 334 train components are manually annotated with bounding boxes. Based on this data set, we carry out extensive experiments that have demonstrated our proposed HGR outperforms the existing state-of-the-art baselines significantly. The data set and the source code can be downloaded online at https://github.com/ChengZY/HGR.

We develop a systematic theory to reconstruct missing samples in a time series using a spatiotemporal memory based on artificial neural networks. The Markov order of the input process is learned and subsequently used for learning temporal correlations from data difference sequences. We enforce the Lipschitz continuity criterion in our algorithm, leading to a regularized optimization framework for learning. The performance of the algorithm is analyzed using both theory and simulations. The efficacy of the technique is tested on synthetic and real life data sets. Our technique is analytic and uses nonlinear feedback within an optimization setup. Simulation results show that the algorithm presented in this article significantly outperforms the state-of-the-art algorithms for missing samples reconstruction with the same data set and similar training conditions.

Design of bridge piers and abutments is significantly impacted by hydrodynamic processes that cause scouring of the foundation. Although, many empirical formulae are available in the literature to estimate the depth of scouring, but they suffer from several limitations. A major limitation of empirical formulae is that they are largely applicable to the hydraulic conditions for which they have been derived. In this research, a deep neural network (DNN) has been developed and applied to predict the depth of scour around bridge piers and abutments. The practicality of the proposed model has been demonstrated using the experimental data sets consisting of 211 data points. The novelty of the DNN model applied herein lies in the use of Adam Optimizer for optimizing the parameters of the DNN model. The performance of the DNN model was evaluated for each parameter set using statistical indicators such as the coefficient of determination, root mean square error, and mean absolute error. A regression equation based upon the available data set has also been proposed. Based upon the values of the statistical parameters, the DNN model has been found to be significantly better than the regression model. The model proposed herein performs better than the regression model. A distinct practical advantage of the model proposed herein is that it eliminates the need of hit and trial procedure to determine the optimal parameter set for the model.

Bearing is one of the critical components in rotating equipment. Therefore, accurate estimation of the remaining useful life (RUL) of bearings plays a vital role in reducing the costly unplanned maintenance and increasing the reliability of machines. This paper proposes a method for bearing prognostics that uses iteratively updated degradation regression models to capture the degradation trend in bearing’s health indicator (HI), and the models are utilized to predict the degradation trajectory of HI and to estimate the RUL of bearings. The importance of determining the time to start prediction by elbow point is explained, which is often overlooked in prognostics. To improve the prognostic performance, an adaptive approach for elbow point detection is designed based on the gradient change of HIs, and a new smooth approach is applied to reduce spurious fluctuations in degradation trajectory. The effectiveness of the proposed method is validated on two publicly available data sets, i.e., IMS and FEMTO bearing prognostics data set, and its prognostic performance is compared with that of three state-of-the-art methods. The obtained results demonstrate that the proposed method can effectively detect elbow point and determine the time to start prediction, and can calibrate the degradation regression model dynamically according to the evolving degradation trend in the HI, which validates its superior prognostic performance.

Capacitated Vehicle routing problem is NP-hard scheduling problem in which the main concern is to find the best routes with minimum cost for a number of vehicles serving a number of scattered customers under some vehicle capacity constraint. Due to the complex nature of the capacitated vehicle routing problem, metaheuristic optimization algorithms are widely used for tackling this type of challenge. Coronavirus Herd Immunity Optimizer (CHIO) is a recent metaheuristic population-based algorithm that mimics the COVID-19 herd immunity treatment strategy. In this paper, CHIO is modified for capacitated vehicle routing problem. The modifications for CHIO are accomplished by modifying its operators to preserve the solution feasibility for this type of vehicle routing problems. To evaluate the modified CHIO, two sets of data sets are used: the first data set has ten Synthetic CVRP models while the second is an ABEFMP data set which has 27 instances with different models. Moreover, the results achieved by modified CHIO are compared against the results of other 13 well-regarded algorithms. For the first data set, the modified CHIO is able to gain the same results as the other comparative methods in two out of ten instances and acceptable results in the rest. For the second and the more complicated data sets, the modified CHIO is able to achieve very competitive results and ranked the first for 8 instances out of 27. In a nutshell, the modified CHIO is able to efficiently solve the capacitated vehicle routing problem and can be utilized for other routing problems in the future such as multiple travelling salesman problem.

The normal growth of fishes is closely relevant to the density of mariculture. It is of great significance to accurately calculate the breeding area of specific sea area from satellite remote sensing images. However, there are no reports about cage segmentation and density detection based on remote sensing images so far. And the accurate segmentation of cages faces challenges from very large high-resolution images. Firstly, a new public mariculture cage data set is built. Secondly, the training set is augmented via sample variations to improve the robustness of the model. Then, for cage segmentation and density statistics, a new methodology based on Mask R-CNN is proposed. Using dividing and stitching technologies, the entire remote sensing test images of the cage can be accurately segmented. Finally, using the trained model, the object detection features and segmentation characteristics can be obtained at the same time. Considering only the area within the target detection frame, the proposed method can count the pixels in the segmented area, which can obtain accurate area and density while reducing time-consuming. Experimental results demonstrate that, compared with traditional contour extraction method and U-Net based scheme, the proposed scheme can significantly improve segmentation precision and model’s robustness. The relative error of the actual area is only 1.3%.

Clustering taxi data is commonly used to understand spatial patterns of urban mobility. In this paper, we propose a new clustering model called Origin-Destination-means (OD-means). OD-means is a hierarchical adaptive k-means algorithm based on origin-destination pairs. In the first layer of the hierarchy, the clusters are separated automatically based on the variation of the within-cluster distance of each cluster until convergence. The second layer of the hierarchy corresponds to the sub clustering process of small clusters based on the distance between the origin and destination of each cluster. The algorithm is tested on a large data set of taxi GPS data from Santiago, Chile, and compared to other clustering algorithms. In contrast to them, our proposed model is capable of detecting general and local travel patterns in the city due to its hierarchical structure.

The ever-increasing popularity of recommendation systems allows users to find appropriate services without excessive effort. However, due to the unstable and complex network environment, the historical behavior data of users are quite sparse in most cases. The inherent drawbacks render preference prediction infeasible for cold-start users and have become a crucial issue to be resolved in recommendation systems. To deal with the problems, we first present a Trust-based Collaborative Filtering (TbCF) algorithm to perform basic rating prediction in a manner consistent with the existing CF methods. Then, we propose the Hybrid Collaborative Filtering Recommendation approach with User-Item-Trust Records (UIThybrid), a novel approach that incorporates user trust into the existing CF-based methods in a harmonious way to supplement rating information. UIThybrid employs multiple perspectives to extract proper services and achieves a good tradeoff between the robustness, accuracy, and diversity of the recommendation. We conduct extensive real-world experiments on the Epinions data set to demonstrate the feasibility and efficiency of UIThybrid.

The meaningful patterns embedded in high-dimensional multi-view data sets typically tend to have a much more compact representation that often lies close to a low-dimensional manifold. Identification of hidden structures in such data mainly depends on the proper modeling of the geometry of low-dimensional manifolds. In this regard, this article presents a manifold optimization-based integrative clustering algorithm for multi-view data. To identify consensus clusters, the algorithm constructs a joint graph Laplacian that contains denoised cluster information of the individual views. It optimizes a joint clustering objective while reducing the disagreement between the cluster structures conveyed by the joint and individual views. The optimization is performed alternatively over k -means and Stiefel manifolds. The Stiefel manifold helps to model the nonlinearities and differential clusters within the individual views, whereas k -means manifold tries to elucidate the best-fit joint cluster structure of the data. A gradient-based movement is performed separately on the manifold of each view so that individual nonlinearity is preserved while looking for shared cluster information. The convergence of the proposed algorithm is established over the manifold and asymptotic convergence bound is obtained to quantify theoretically how fast the sequence of iterates generated by the algorithm converges to an optimal solution. The integrative clustering on benchmark and multi-omics cancer data sets demonstrates that the proposed algorithm outperforms state-of-the-art multi-view clustering approaches.

This paper considers a problem of estimating brain effective connectivity from EEG signals using a Granger causality (GC) concept characterized on state-space models. We propose a state-space model for explaining coupled dynamics of the source and EEG signals where EEG is a linear combination of sources according to the characteristics of volume conduction. Our formulation has a sparsity prior on the source output matrix that can further classify active and inactive sources. The scheme is comprised of two main steps: model estimation and model inference to estimate brain connectivity. The model estimation consists of performing a subspace identification and the active source selection based on a group-norm regularized least-squares. The model inference relies on the concept of state-space GC that requires solving a discrete-time Riccati equation for the covariance of estimation error. We verify the performance on simulated data sets that represent realistic human brain activities under several conditions including percentages of active sources, a number of EEG electrodes and the location of active sources. The performance of estimating brain networks is compared with a two-stage approach using source reconstruction algorithms and VAR-based Granger analysis. Our method achieved better performances than the two-stage approach under the assumptions that the true source dynamics are sparse and generated from state-space models. The method is applied to a real EEG SSVEP data set and we found that the temporal lobe played a role of a mediator of connections between temporal and occipital areas, which agreed with findings in previous studies.

The prediction of lane changes can reduce traffic accidents and improve traffic flow. In this paper two classifiers, Echo State Network and a recurrent neural network with Long Short Term Memory cells, were compared to predict lane changes using the input variables steering angle and indicator. The input variables were extracted from a data set which was generated from a naturalistic driving study in the urban area of Chemnitz, Germany. Both classifiers predicted left and right lane changes successfully. They achieved high true positive rates and low false positive rates. The Echo State Network predicted left and the recurrent neural network predicted right lane changes better.

Communication constraints are one of the major challenges preventing the wide-spread adoption of Federated Learning systems. Recently, Federated Distillation (FD), a new algorithmic paradigm for Federated Learning with fundamentally different communication properties, emerged. FD methods leverage ensemble distillation techniques and exchange model outputs, presented as soft labels on an unlabeled public data set, between the central server and the participating clients. In this work, we investigate FD from the perspective of communication efficiency by analyzing the effects of active distillation-data curation, soft-label quantization, and delta-coding techniques. Based on the insights gathered from this analysis, we present Compressed Federated Distillation (CFD), an efficient Federated Distillation method. Extensive experiments, on federated image classification and language modeling problems, at different levels of data heterogeneity, demonstrate that our method can reduce the amount of communication necessary to achieve fixed performance targets by more than two orders of magnitude when compared to FD, and by more than four orders of magnitude when compared to parameter averaging based techniques like Federated Averaging.

Mathematical morphology spectrum entropy is a signal feature extraction method based on information entropy and mathematical morphology. The scale of structure element is a critical parameter, whose value determines the accuracy of feature extraction. Existing scale selection methods depend on experiment parameters or external indicators including noise ratio, fault frequencies, etc. In many cases, existing methods obtain fix scale and they are not suitable for quantifying the performance degradation and the fault degree of bearings. There are few researches on scale selection based on the properties of mathematical morphology spectrum. In this study, a scale-adaptive mathematical morphology spectrum entropy (AMMSE) is proposed to improve the scale selection. To support the proposed method, two properties of the mathematical morphology spectrum (MMS), namely non-negativity and monotonic decreasing, are proved. It can be concluded from the two properties that the feature loss of MMS decreases with the increase of scale. Based on the conclusion, two adaptive scale selection strategies are proposed to automatically determine the scale by reducing the feature loss of MMS. AMMSE is the integration of two strategies. Compare to the existing methods, AMMSE is not constrained by the information of the experiment and the signal. The scale of AMMSE changes with the signal characteristics and is no longer fixed by experimental parameters. The parameters of AMMSE are more generalizable as well. The presented method is applied to identify fault degree on CWRU bearing data set and evaluate performance degradation on IMS bearing data set. The experiment result shows that AMMSE has better results in both experiments with the same parameters.

International audience; K-anonymity is the most widely used approach to privacy preserving microdata which is mainly based on generalization. Although generalization-based k-anonymity approaches can achieve the privacy protection objective, they suffer from information loss. Clustering-based approaches have been successfully adapted for k-anonymization as they enhance the data quality, however, the computational complexity of finding an optimal solution has shown as NP-hard. Nature-inspired optimization algorithms are effective in finding solutions to complex problems. We propose, in this paper, a novel algorithm based on a simple nature-inspired metaheuristic called Black Hole Algorithm (BHA), to address such limitations. Experiments on real data set show that data utility has been improved by our approach compared to k-anonymity, BHA-based k-anonymity and clustering-based k-anonymity approaches.

Imbalanced class distribution is an inherent problem in many real-world classification tasks where the minority class is the class of interest. Many conventional statistical and machine learning classification algorithms are subject to frequency bias, and learning discriminating boundaries between the minority and majority classes could be challenging. To address the class distribution imbalance in deep learning, we propose a class rebalancing strategy based on a class-balanced dynamically weighted loss function where weights are assigned based on the class frequency and predicted probability of ground-truth class. The ability of dynamic weighting scheme to self-adapt its weights depending on the prediction scores allows the model to adjust for instances with varying levels of difficulty resulting in gradient updates driven by hard minority class samples. We further show that the proposed loss function is classification calibrated. Experiments conducted on highly imbalanced data across different applications of cyber intrusion detection (CICIDS2017 data set) and medical imaging (ISIC2019 data set) show robust generalization. Theoretical results supported by superior empirical performance provide justification for the validity of the proposed dynamically weighted balanced (DWB) loss function.

Attribute reduction, also called feature selection, is one of the most important issues of rough set theory, which is regarded as a vital preprocessing step in pattern recognition, machine learning, and data mining. Nowadays, high-dimensional mixed and incomplete data sets are very common in real-world applications. Certainly, the selection of a promising feature subset from such data sets is a very interesting, but challenging problem. Almost all of the existing methods generated a cover on the space of objects to determine important features. However, some tolerance classes in the cover are useless for the computational process. Thus, this article introduces a new concept of stripped neighborhood covers to reduce unnecessary tolerance classes from the original cover. Based on the proposed stripped neighborhood cover, we define a new reduct in mixed and incomplete decision tables, and then design an efficient heuristic algorithm to find this reduct. For each loop in the main loop of the proposed algorithm, we use an error measure to select an optimal feature and put it into the selected feature subset. Besides, to deal more efficiently with high-dimensional data sets, we also determine redundant features after each loop and remove them from the candidate feature subset. For the purpose of verifying the performance of the proposed algorithm, we carry out experiments on data sets downloaded from public data sources to compare with existing state-of-the-art algorithms. Experimental results showed that our algorithm outperforms compared algorithms, especially in classification accuracy.

In the process industry, it is essential to establish a data-driven soft sensor to predict the key variable that is difficult to online measure directly. The accuracy performance of data-driven soft sensors relies heavily on data. Unfortunately, it is hard to acquire sufficient and informative data from the samples with limited number, which is called as the small sample problem. For handling the small sample problem, it is a good solution to generating virtual samples according to the distribution of original data. This paper proposes an enhanced method of virtual sample generation utilizing manifold features to develop soft sensors using small data. First, T-Distribution Stochastic Neighbor Embedding (t-SNE) is utilized to extract the features of input data. The main idea of generating virtual samples is to use the interpolation algorithm to obtain virtual t-SNE input features and then the random forest algorithm is utilized to get the virtual outputs using virtual t-SNE input features. Finally, virtual samples using the proposed t-SNE based virtual sample generation (t-SNE-VSG) can be achieved. For the sake of confirming the effectiveness and feasibility of the presented t-SNE-VSG, a standard data set is first used. What is more, a small data set from an actual industrial process of Purified Terephthalic Acid is used to establish a soft sensor model. The results from simulations show that the accuracy performance of the soft sensor established with small data can be effectively improved by adding the virtual samples generated by t-SNE-VSG. In addition, t-SNE-VSG achieves superior accuracy to state-of-the-art virtual sample generation methods.

The smart unmanned vending machine using machine vision technology suffers from the sharp decrease of detection accuracy due to the incomplete image collection of items by monocular camera in complex environment, and the lack of obvious features in dense stacking of items. In this paper, a binocular camera system is designed to effectively solve the problems of distortion and coverage caused by monocular camera. Besides, an image-stitching algorithm is developed to splice the images captured by the camera, which reliefs the burden of computation for back-end recognition processing brought by the binocular camera. A new model YOLOv3-TinyE is proposed based on YOLOv3-tiny model. Based on the data set of 21, 000 images captured in real scenarios that containing 20 different type of beverages, the comparison experimental results show that YOLOv3-TinyE model achieves the mean average precision of 99.15%, and the inference speed is 2.91 times faster than that of YOLOv3 model, and the detection accuracy of YOLOv3-TinyE model based on binocular vision is higher than that based on monocular vision. The results suggest that the designed method achieves the goal in terms of inference speed and average precision, that is, it is able to satisfy the requirements for real-world applications.

MixUp is an effective data augmentation method to regularize deep neural networks via random linear interpolations between pairs of samples and their labels. It plays an important role in model regularization, semisupervised learning (SSL), and domain adaption. However, despite its empirical success, its deficiency of randomly mixing samples has poorly been studied. Since deep networks are capable of memorizing the entire data set, the corrupted samples generated by vanilla MixUp with a badly chosen interpolation policy will degrade the performance of networks. To overcome overfitting to corrupted samples, inspired by metalearning (learning to learn), we propose a novel technique of learning to a mixup in this work, namely, MetaMixUp. Unlike the vanilla MixUp that samples interpolation policy from a predefined distribution, this article introduces a metalearning-based online optimization approach to dynamically learn the interpolation policy in a data-adaptive way (learning to learn better). The validation set performance via metalearning captures the noisy degree, which provides optimal directions for interpolation policy learning. Furthermore, we adapt our method for pseudolabel-based SSL along with a refined pseudolabeling strategy. In our experiments, our method achieves better performance than vanilla MixUp and its variants under SL configuration. In particular, extensive experiments show that our MetaMixUp adapted SSL greatly outperforms MixUp and many state-of-the-art methods on CIFAR-10 and SVHN benchmarks under the SSL configuration.

The paper proposes a technique for Human Action Recognition (HAR) that uses a Convolutional Neural Network (CNN). Depth data sequences from the motion sensing devices are converted into images and fed into a CNN rather than using any conventional or statistical method. The initial data was obtained from 10 actions performed by six subjects captured by the Kinect v2 sensor as well as 20 actions performed by 7 subjects from the MSR 3D Action data set. A custom CNN architecture consisting of three convolutional and three max pooling layers followed by a fully connected layer was used. Training, validation, and testing was carried out on a total of 39715 images. An accuracy of 97.23% was achieved on the Kinect data set. On the MSR data set the accuracy was 87.1%.

We propose a new approach for estimating rare disaster event models where we only use U.S. national consumption data as an alternative to the ubiquitous Barro and Urs´ua’s (2008, 2012) multi-country data set. We find that the 2020 COVID crisis unambiguously reveals the presence and significance of rare disasters in consumption dynamics. Using our estimated parameters and recursive preferences, our approach is able to solve the risk-free rate and equity premium puzzles without resorting to multi-country data in estimating the model. Our analysis shows that the severity of disasters is vastly underestimated in the U.S. and that more than 200 years of consumption data would be necessary to get an accurate estimate.

Partial discharge (PD) monitoring of electrical substations could provide early warning of insulation failures. Among the various technologies, Radio Frequency (RF) based PD monitoring system could be a promising solution. The RF-based monitoring system detects PD sources in the substation and can also localise the PD sources. The time difference of arrival (TDOA) based PD localisation system primarily require arrival time of the impulsive RF signal. Though many localisation algorithms have been proposed in the recent past to overcome the TDOA estimation errors, less attention has been given to the accurate estimation of RF PD signal arrival time. This paper presents the AT's automatic labelling in the RF PD signal using Bi-Long Short-Term Memory (Bi-LSTM) network applied on the continuous wavelet transformed (CWT) signal. Further, it also shows PD signal augmentation to overcome the problem of limited representative training data set. The behaviour of the radiated RF signals is influenced by many factors and has almost stochastic characteristics. The proposed system has been validated with laboratory-based experimental signals and the data set obtained from different electrical substations. The results show that the improved performance is obtained from the combination of a multilayer Bi-LSTM model and an augmented training set.

Many methods have been proposed to estimate how much effort is required to build and maintain software. Much of that research tries to recommend a single method - an approach that makes the dubious assumption that one method can handle the diversity of software project data. To address that drawback, we apply a configuration technique called “ROME” (Rapid Optimizing Methods for Estimation), which uses sequential model-based optimization (SMO) to find what configuration settings of effort estimation techniques work best for a particular data set. We test this method using data from 1161 classic waterfall projects and 120 contemporary projects (from GitHub). In terms of magnitude of relative error and standardized accuracy, we find that ROME achieves better performance than the state-of-the-art methods for both classic and contemporary projects. In addition, we conclude that we should not recommend one method for estimation. Rather, it is better to search through a wide range of different methods to find what works best for the local data. To the best of our knowledge, this is the largest effort estimation experiment yet attempted and the only one to test its methods on classic and contemporary projects.

The KNN classification algorithm is one of the most commonly used algorithm in the AI field. This paper proposes two improved algorithms, namely KNNTS, and KNNTS-PK+. The two improved algorithms are based on KNNPK+ algorithm, which uses PK-Means ++ algorithm to select the center of the spherical region, and sets the radius of the region to form a sphere to divide the data set in the space. The KNNPK+ algorithm improves the classification accuracy on the premise of stabilizing the classification efficiency of KNN classification algorithm. In order to improve the classification efficiency of KNN algorithm on the premise that the accuracy of KNN classification algorithm remains unchanged, KNNTS algorithm is proposed. It uses tabu search algorithm to select the radius of spherical region, and uses spherical region division method with equal radius to divide the data set in space. On the basis of the first two improved algorithms, KNNTS-PK+ algorithm combines them to divide the data sets in space. Experiments are carried out on the new data set and the classification results were obtained. Results revealed show that the two improved algorithms can effectively improve the classification accuracy and efficiency after the data samples are cut reasonably.

We present an introspective failure prediction approach for autonomous vehicles. In autonomous driving, complex or unknown scenarios can cause a disengagement of the self-driving system. Disengagements can be triggered either by automatic safety measures or by human intervention. We propose to use recorded disengagement sequences from test drives as training data to learn to predict future failures. The system then learns introspectively from its own previous mistakes. In order to predict failures as early as possible, we propose a machine learning approach where sequences of sensor data are classified as either failure or success. The car itself is treated as a black box. Our method combines two sensor modalities that contain different types of information. An image-based model learns to detect generally challenging situations such as crowded intersections accurately multiple seconds in advance. A state data based model allows to detect fast changes immediately before a failure, such as sudden braking or swerving. The outcome of the individual models is fused by averaging the individual failure probabilities. We evaluate our approach on a data set provided by the BMW Group containing 14 hours of autonomous driving. The proposed late fusion approach allows for predicting failures at an accuracy of more than 85% seven seconds in advance, at a false positive rate of 20%. The proposed method outperforms state-of-the-art failure prediction by more than 15% while being a flexible framework that allows for straightforward addition of further sensor modalities.

The goal of this study is to provide a framework, using hidden semi-Markov models, for modeling a driver's response time after an alert is provided in manual driving. Given the plethora of alerts and warning within a vehicle, there is a need to understand when a driver will respond after an alert is provided. Data from a previous driving simulator study, where drivers were interacting with an in-vehicle information system (IVIS) were used for model training. The final data set included 16 participants, with 288 task initiations. The proposed model could predict a driver's response time accurately using only a small portion of the available data, and had a mean absolute error of 0.51 seconds with 84% of predictions within an absolute error of 1 second. This framework has applicability in mitigating the risk of transitions in driver distraction. This includes transitions from the road to the secondary task and back to the road.

Pattern mining is a fundamental technique of data mining to discover interesting correlations in the data set. There are several variations of pattern mining, such as frequent itemset mining, sequence mining, and high utility itemset mining. High utility itemset mining is an emerging data science task, aims to extract knowledge based on a domain objective. The utility of a pattern shows its effectiveness or benefit that can be calculated based on user priority and domain-specific understanding. The sequential pattern mining (SPM) issue is much examined and expanded in various directions. Sequential pattern mining enumerates sequential patterns in a sequence data collection. Researchers have paid more attention in recent years to frequent pattern mining over uncertain transaction dataset. In recent years, mining itemsets in big data have received extensive attention based on the Apache Hadoop and Spark framework. This paper seeks to give a broad overview of the distinct approaches to pattern mining in the Big Data domain. Initially, we investigate the problem involved with pattern mining approaches and associated techniques such as Apache Hadoop, Apache Spark, parallel and distributed processing. Then we examine major developments in parallel, distributed, and scalable pattern mining, analyze them in the big data perspective and identify difficulties in designing the algorithms. In particular, we study four varieties of itemsets mining, i.e., parallel frequent itemsets mining, high utility itemset mining, sequential patterns mining and frequent itemset mining in uncertain data. This paper concludes with a discussion of open issues and opportunity. It also provides direction for further enhancement of existing approaches.

In this paper, we propose a maximum-a-posteriori probability (MAP) based velocity estimation technique in which the prior distribution is defined by current location of the user. Motivation of this work is to improve accuracy of the existing velocity estimation techniques which are either solely based on cellular network measurements or location specific information. Our objective is to exploit both cellular measurements and location information in Bayesian sense; thus, to jointly address the critical applications of mobility management in Heterogeneous-Networks (HetNets), and intelligent transportation system. Here we assume that the Next Generation Simulation (NGSIM) data set for velocity is available at the current location and can be utilized to approximate the prior distribution. Additional information in form of prior distribution function is then exploited to improve the minimum variance unbiased (MVU) estimate of velocity which is based on handover count measurements. Since MVU estimate is a random variable, we first formulate its density function parameterized over the actual velocity. Next, we follow Bayesian approach to accommodate both prior distribution and parametric density function in deriving posterior density function of velocity. Finally, we derive expression of the MAP estimator considering various standard distribution functions which best fit to the density function obtained from NGSIM data set. In order to quantify the quality of estimate, we derive its variance and the corresponding Cramer-Rao-bound (CRB) on the minimum error variance. Numerical results demonstrate that the proposed estimator which incorporates NGSIM data set is asymptotically efficient and outperforms other classical handover count based estimation techniques.

Predictive process monitoring aims to produce early warnings of unwanted events. We consider the use of the machine learning method extreme gradient boosting as the forecasting model in predictive monitoring. A tuning algorithm is proposed as the signaling method to produce a required false alarm rate. We demonstrate the procedure using a unique data set on mental health in the Netherlands. The goal of this application is to support healthcare workers in identifying the risk of a mental health crisis in people diagnosed with schizophrenia. The procedure we outline offers promising results and a novel approach to predictive monitoring.

Brain computer interface (BCI) P300 speller can help severely disabled patients communicate and control with external machines or robots, so that the classification methods of P300 Electroencephalogram (EEG) signal play an important role in the development of BCI system and technology. In this paper, a novel support vector neural network (SVNN) is proposed and developed to obtain more accurate and effective EEG classification results. It is the first time to combine linear variational inequality (LVI) based primal-dual neural network with convex quadratic programming problem based on support vector machine to solve the classification problem. It has been proved that the support vector neural network globally converges to the optimal solution of convex optimization problem and iterates the parameters in the form of matrix, which means that the method has global convergence and parallelism. The proposed SVNN method is used to solve the classification problem of P300 EEG signals. Experimental results on data set IIb from BCI competition II and data set II from BCI competition III show that the accuracy of the proposed SVNN method is 100% and 98%, respectively. Compared with most of the state-of-theart algorithms, SVNN has the highest recognition accuracy and information transfer rate.

Touch modality identification has attracted increasing attention due to its importance in human-robot interactions. There are three issues involved in the tactile perception for the touch modality identification, including the high dimensionality of tactile signals, complex tensor morphology of tactile sensing units, and the misalignment among different tactile time-series samples. In this article, we propose a novel kernel-based approach to deal with these three issues in a unified framework. Specifically, the techniques, including sparse principal component analysis and subsampling, are employed to reduce the feature dimension. Then, a singular value decomposition (SVD)-based kernel is proposed to preserve the spatial information of the tactile sensing elements. The sample misalignment issue is addressed via the employment of a global alignment kernel. Moreover, the merits of these two kernels are fused through an ideal regularized composite kernel, which simultaneously takes the label information of the training set into consideration. The effectiveness of the proposed kernel-based approach is verified on a public touch modality data set with a comprehensive comparison with the competing methods.

At present, there is a problem of insufficient utilization of regional features in the application of the existing knowledge graph based on GGNN in the defect classification of the bolt and nut pair of transmission lines. Therefore, a decision-making method of combining the bolt and nut pair with the original regional feature and the bolt and nut pair knowledge graph feature is proposed. For this reason, a method is proposed to combine the decision-making of the bolt and nut pair on the original regional features and the bolt and nut pair on the characteristics of the knowledge graph. First, the characteristics of the bolt and nut pair knowledge graph are combined with the adaptive normalization of the bolt and nut pair to the features of the original area. Then, the classification score vector based on fusion features and the classification score vector based on the bolt and nut pair to the original area feature are derived from the classifier respectively; Finally, the classification score vector of the fusion feature and the bolt and nut pair are fused to the classification score vector of the original region feature to obtain the final classification result. On this basis, this article uses bolt and nut pair to conduct multiple sets of defect classification experiments on the data set of the knowledge graph experiment. The experimental results show that the method of decision fusion of the bolt and nut pair to the original regional feature fusion the bolt and nut pair to the knowledge graph feature is better than the bolt and nut pair to the knowledge graph average precision, precision, and recall rate. It is effective Prove the improvement of the algorithm.

Spectral clustering is becoming more and more popular because it has good performance in discovering clusters with varying characteristics. However, it suffers from high computational cost, unstable clustering results and noises. This work presents a novel approximate spectral clustering based on dense cores and density peaks, called DCDP-ASC. It first finds a reduced data set by introducing the concept of dense cores; then defines a new distance based on the common neighborhood of dense cores and calculates geodesic distances between dense cores according to the new defined distance; after that constructs a decision graph with a parameter-free local density and geodesic distance for obtaining initial centers; finally calculates the similarity between dense cores with their new defined geodesic distance, employs normalized spectral clustering method to divide dense cores, and expands the result on dense cores to the whole data set by assigning each point to its representative. The results on some challenging data sets and the comparison of our algorithm with some other excellent methods demonstrate that the proposed method DCDP-ASC is more advantageous in identifying complex structured clusters containing a lot of noises.

Optical coherence tomography (OCT) has been identified as a non-invasive and inexpensive imaging modality to discover potential biomarkers for Alzheimer's diagnosis and progress determination. Current hypotheses presume the thickness of the retinal layers, which are analyzable within OCT scans, as an effective biomarker for the presence of Alzheimer's. As a logical first step, this work concentrates on the accurate segmentation of retinal layers to isolate the layers for further analysis. This paper proposes a generative adversarial network (GAN) that concurrently learns to increase the image resolution for higher clarity and then segment the retinal layers. We propose a multi-stage and multi-discriminatory generative adversarial network (MultiSDGAN) specifically for superresolution and segmentation of OCT scans of the retinal layer. The resulting generator is adversarially trained against multiple discriminator networks at multiple stages. We aim to avoid early saturation of generator model training leading to poor segmentation accuracies and enhance the process of OCT domain translation by satisfying all the discriminators in multiple scales. We also investigated incorporating the Dice loss and Structured Similarity Index Measure (SSIM) as additional loss functions to specifically target and improve our proposed GAN architecture's segmentation and superresolution performance, respectively. The ablation study results conducted on our data set suggest that the proposed MultiSDGAN with ten-fold cross-validation (10-CV) provides a reduced equal error rate with 44.24% and 34.09% relative improvements, respectively (p-values of the improvement level tests .01). Furthermore, our experimental results also demonstrate that the addition of the new terms to the loss function improves the segmentation results significantly by relative improvements of 31.33% (p-value .01).

This article introduces a novel shallow 3-D self-supervised tensor neural network in quantum formalism for volumetric segmentation of medical images with merits of obviating training and supervision. The proposed network is referred to as the 3-D quantum-inspired self-supervised tensor neural network (3-D-QNet). The underlying architecture of 3-D-QNet is composed of a trinity of volumetric layers, viz., input, intermediate, and output layers interconnected using an S -connected third-order neighborhood-based topology for voxelwise processing of 3-D medical image data, suitable for semantic segmentation. Each of the volumetric layers contains quantum neurons designated by qubits or quantum bits. The incorporation of tensor decomposition in quantum formalism leads to faster convergence of network operations to preclude the inherent slow convergence problems faced by the classical supervised and self-supervised networks. The segmented volumes are obtained once the network converges. The suggested 3-D-QNet is tailored and tested on the BRATS 2019 Brain MR image dataset and the Liver Tumor Segmentation Challenge (LiTS17) dataset extensively in our experiments. The 3-D-QNet has achieved promising dice similarity (DS) as compared with the time-intensive supervised convolutional neural network (CNN)-based models, such as 3-D-UNet, voxelwise residual network (VoxResNet), Dense-Res-Inception Net (DRINet), and 3-D-ESPNet, thereby showing a potential advantage of our self-supervised shallow network on facilitating semantic segmentation.

Positron Emission Tomography (PET) scanners are usually designed with the goal to obtain the best compromise between sensitivity, resolution, field-of-view size, and cost. Therefore, it is difficult to improve the resolution of a PET scanner with hardware modifications, without affecting some of the other important parameters. Iterative image reconstruction methods such as the ordered subsets expectation maximization (OSEM) algorithm are able to obtain some resolution recovery by using a realistic system response matrix that includes all the relevant physical effects. Nevertheless, this resolution recovery is often limited by reduced sampling in the projection space, determined by the geometry of the detector. The goal of this work is to improve the resolution beyond the detector size limit by increasing the sampling with data-driven interpolated data. A maximum-likelihood estimation of the counts in each virtual sub-line-of-response (subLOR) is obtained after a complete image reconstruction, conserving the statistics of the initial data set. The new estimation is used for the next complete reconstruction. The method typically requires two or three of these full reconstructions (superiterations). We have evaluated it with simulations and real acquisitions for the Argus and Super Argus preclinical PET scanners manufactured by SMI, considering different types of increased sampling. Quantitative measurements of recovery and resolution evolution against noise per iteration for the standard OSEM and successive superiterations show promising results. The procedure is able to reduce significantly the impact of depth-of-interaction in large crystals, and to improve the spatial resolution. The proposed method is quite general and it can be applied to other scanners and configurations., Comment: 5 pages, 8 figures, conference proceedings

Federated learning provides a promising paradigm to enable network edge intelligence in the future sixth generation (6G) systems. However, due to the high dynamics of wireless circumstances and user behavior, the collected training data is non-independent and identically distributed (non-IID), which causes severe performance degradation of federated learning. To solve this problem, federated learning with non-IID data in wireless networks is studied in this paper. Firstly, based on the derived upper bound of expected weight divergence, a federated averaging scheme is proposed to reduce the distribution divergence of non-IID data. Secondly, to further harmonize the distribution divergence, data sharing is associated with federated learning in wireless networks, and a joint optimization algorithm is designed to keep a sophisticated balance between the model accuracy and the cost. Finally, the simulation results based on a common-used image data set are provided to evaluate the performance of our proposed schemes, which can achieve significant performance gains with a small price of latency and energy consumption.