Search

Rumors can cause devastating consequences to individuals and our society. Analysis shows that the widespread of rumors typically results from deliberate promotion of information aiming to shape the collective public opinions on the concerned event. In this paper, we combat such chaotic phenomenon with a countermeasure by mirroring against how such chaos is created to make rumor detection more robust and effective. Our idea is inspired by adversarial learning method originated from Generative Adversarial Networks (GAN). We propose a GAN-style approach, where a generator is designed to produce uncertain or conflicting voices, further polarizing the original conversational threads to boost the discriminator. We reveal that feature learning effectiveness is highly relevant to the quality of generated parody. Given the strong natural language generation performance of transformer, we propose a transformer-based method to improve the generated posts, which appear to be closely responsive to the source post and retain the authentic propagation structure. Different from traditional data-driven rumor detection approaches, our method can capture low-frequency but more salient non-trivial discriminant patterns. Extensive experiments on THREE benchmarks demonstrate that our rumor detection method achieves much better results than state-of-the-art methods, and the transformer-based model further improve the performance of our GAN-style approach.

With the development of science and technology, the rapid development of social economy, the motor as a new type of transmission equipment, in the production and life of people occupies a pivotal position. Under the rapid development of computer and electronic technology, manufacturing equipment is becoming larger, faster, more continuous, and more automated. This has resulted in complex, expensive, accident-damaging, and high-impact equipment for electric motors; even routine maintenance requires significant equipment maintenance and maintenance costs. If a fault occurs, it will cause serious damage to the entire equipment and can even have a major impact on the entire production process, leading to a serious economic and social life. In this paper, a CNN-based machine learning fault diagnosis method is proposed to address the problem of high incidence of motor faults and difficulty in identifying fault types. A fault reproduction test is constructed by machine learning techniques to extract vibration time domain data for normal operating conditions, rotor eccentricity, stator short circuit, and bearing inner ring fault; divide the data segment into 15 speed segments, extract 13 typical time domain features for each speed segment; and perform mathematical statistics for fault diagnosis. Compared with the traditional algorithm, the method has more comprehensive feature information extraction, higher diagnostic accuracy, and faster diagnostic speed, with a fault diagnosis accuracy of 98.7%.

International audience; Advanced technologies for the calibration of machine tool are presented. Geometric, kinematic, thermally-induced and load-induced errors are classified into errors within one-axis as intra-axis errors, errors between axes as inter-axis errors, and as volumetric errors. As the major technological elements of machine tool calibration, the measurement methods, mathematical models and compensation approaches of the machine tool errors are addressed. The criteria for selecting a combination of the technological elements for machine tool calibration from the point of view of accuracy, complexity, and cost are provided. Recent applications of artificial intelligence and machine learning in machine tool calibration are introduced. Remarks are also made on future trends in machine tool calibration.

Emergency response activity relies on transportation networks. Emergency facility location interacts with transportation networks clearly. This review is aimed to provide a combined framework for emergency facility location in transportation networks. The article reveals emergency response activities research clusters, issues, and objectives according to keywords co-occurrence analysis. Four classes of spatial separation models in transportation networks, including distance, routing, accessibility, and travel time are introduced. The stochastic and time-dependent characteristics of travel time are described. Travel time estimation and prediction method, travel time under emergency vehicle preemption, transportation network equilibrium method, and travel time in degradable networks are demonstrated. The emergency facilities location models interact with transportation networks, involving location-routing model, location models embedded with accessibility, location models embedded with travel time, and location models employing mathematical program with equilibrium constraints are reviewed. We then point out the-state-of-art challenges: ilities-oriented, evolution landscape and sequential decision modelling, data-driven optimization approach, and machine learning-based algorithms.

A brain-computer interface (BCI) measures and analyzes brain activity and converts this activity into computer commands to control external devices. In contrast to traditional BCIs that require a subject-specific calibration process before being operated, a subject-independent BCI learns a subject-independent model and eliminates subject-specific calibration for new users. However, building subject-independent BCIs remains difficult because electroencephalography (EEG) is highly noisy and varies by subject. In this study, we propose an invariant pattern learning method based on a convolutional neural network (CNN) and big EEG data for subject-independent P300 BCIs. The CNN was trained using EEG data from a large number of subjects, allowing it to extract subject-independent features and make predictions for new users. We collected EEG data from 200 subjects in a P300-based spelling task using two different types of amplifiers. The offline analysis showed that almost all subjects obtained significant cross-subject and cross-amplifier effects, with an average accuracy of more than 80%. Furthermore, more than half of the subjects achieved accuracies above 85%. These results indicated that our method was effective for building a subject-independent P300 BCI, with which more than 50% of users could achieve high accuracies without subject-specific calibration.

Review rating prediction is commonly approached from the perspective of either Collaborative Filtering (CF) or Sentiment Classification (SC). CF-based approach usually resorts to matrix factorization based on user–item interaction, and does not fully utilize the valuable review text features. In contrast, SC-based approach is focused on mining review content, but can just incorporate some user- and product-level features, and fails to capture sufficient interactions between them represented typically in a sparse matrix as CF can do. In this paper, we propose a novel, extensible review rating prediction model called InterSentiment by bridging the user-product interaction model and the sentiment model based on deep learning. InterSentiment is a specific instance of our proposed Deep Learning based Collaborative Filtering framework. The proposed model aims to learn the high-level representations combining user-product interaction and review sentiment, and jointly project them into the rating scores. Results of experiments conducted on IMDB and two Yelp datasets demonstrate clear advantage of our proposed approach over strong baseline methods.

Making an erroneous decision may cause serious results in diverse mission-critical tasks such as medical diagnosis and bioinformatics. Previous work focuses on classification with a reject option, i.e., abstain rather than classify an instance of low confidence. Most mission-critical tasks are always accompanied with class imbalance and cost sensitivity, where AUC has been shown a preferable measure than accuracy in classification. In this work, we propose the framework of AUC optimization with a reject option, and the basic idea is to withhold the decision of ranking a pair of positive and negative instances with a lower cost, rather than mis-ranking. We obtain the Bayes optimal solution for ranking, and learn the reject function and score function for ranking, simultaneously. An online algorithm has been developed for AUC optimization with a reject option, by considering the convex relaxation and plug-in rule. We verify, both theoretically and empirically, the effectiveness of the proposed algorithm.

A novel machine learning aided structural reliability analysis for functionally graded frame structures against static loading is proposed. The uncertain system parameters, which include the material properties, dimensions of structural members, applied loads, as well as the degree of gradation of the functionally graded material (FGM), can be incorporated within a unified structural reliability analysis framework. A 3D finite element method (FEM) for static analysis of bar-type engineering structures involving FGM is presented. By extending the traditional support vector regression (SVR) method, a new kernel-based machine learning technique, namely the extended support vector regression (X-SVR), is proposed for modelling the underpinned relationship between the structural behaviours and the uncertain system inputs. The proposed structural reliability analysis inherits the advantages of the traditional sampling method (i.e., Monte-Carlo Simulation) on providing the information regarding the statistical characteristics (i.e., mean, standard deviations, probability density functions and cumulative distribution functions etc.) of any concerned structural outputs, but with significantly reduced computational efforts. Five numerical examples are investigated to illustrate the accuracy, applicability, and computational efficiency of the proposed computational scheme.

Collaborative learning has gained great popularity due to its benefit of data privacy protection: participants can jointly train a Deep Learning model without sharing their training sets. However, recent works discovered that an adversary can fully recover the sensitive training samples from the shared gradients. Such reconstruction attacks pose severe threats to collaborative learning. Hence, effective mitigation solutions are urgently desired. In this paper, we propose to leverage data augmentation to defeat reconstruction attacks: by preprocessing sensitive images with carefully-selected transformation policies, it becomes infeasible for the adversary to extract any useful information from the corresponding gradients. We design a novel search method to automatically discover qualified policies. We adopt two new metrics to quantify the impacts of transformations on data privacy and model usability, which can significantly accelerate the search speed. Comprehensive evaluations demonstrate that the policies discovered by our method can defeat existing reconstruction attacks in collaborative learning, with high efficiency and negligible impact on the model performance., Comment: 17 pages, 16 figures, CVPR2021 oral

Face recognition is greatly improved by deep convolutional neural networks (CNNs). Recently, these face recognition models have been used for identity authentication in security sensitive applications. However, deep CNNs are vulnerable to adversarial patches, which are physically realizable and stealthy, raising new security concerns on the real-world applications of these models. In this paper, we evaluate the robustness of face recognition models using adversarial patches based on transferability, where the attacker has limited accessibility to the target models. First, we extend the existing transfer-based attack techniques to generate transferable adversarial patches. However, we observe that the transferability is sensitive to initialization and degrades when the perturbation magnitude is large, indicating the overfitting to the substitute models. Second, we propose to regularize the adversarial patches on the low dimensional data manifold. The manifold is represented by generative models pre-trained on legitimate human face images. Using face-like features as adversarial perturbations through optimization on the manifold, we show that the gaps between the responses of substitute models and the target models dramatically decrease, exhibiting a better transferability. Extensive digital world experiments are conducted to demonstrate the superiority of the proposed method in the black-box setting. We apply the proposed method in the physical world as well., Comment: Accpeted by CVPR 2021. Based on the camera ready version, some typos are fixed

Background: Tuberculosis is a major global public health problem. However, it is still in the exploratory stage that the study of the meteorological factors related to the incidence of tuberculosis in Shanxi Province. Therefore, it is very urgent to establish an early warning system that easily operate of tuberculosis. Method: The epidemiological characteristics of tuberculosis in Shanxi Province were described, and the Dynamic Bayesian Network early warning model was established by time series cross-correlation analysis and Bayesian Network.Results: 1. The reported incidence of tuberculosis in Shanxi Province showed an overall downward trend from 2008 to 2017, showing a phenomenon of high in the middle and low at both ends each year, with certain seasonal characteristics. 2. Based on the results of cross-correlation analysis, it is reasonable to use dynamic Bayesian model fitting with meteorological factors lagging for 2 months; the monthly average temperature and monthly precipitation are positively correlated with the incidence of tuberculosis, but the monthly mean air pressure is negatively. 3. Comparison of classification and recognition performance of the three models shows that DBN has the highest classification accuracy in the two regions, which indicates that DBN is better than the other two models in reflecting the performance of minority classes, and better for the comprehensive classification of minority classes and majority classes. Conclusion: 1. Shanxi Province has tuberculosis clustering in time, space and time and space. Incidence peak is in spring and early summer. March is the highest month in the year. Seven meteorological factors such as monthly precipitation are the main factors affecting the incidence of tuberculosis in Shanxi Province. 2. The classification and recognition performance of the Dynamic Bayesian Network early warning model of tuberculosis-meteorological factors established in this study is significantly better than that of static Bayesian Network and support vector machine model, and can better predict the future.

Micro-videos, a new form of videos that are constrained in duration, gain significant popularity in recent years. The volume and rate of online micro-videos urgently calls for effective recommendation algorithms to help users find their interested ones. Although some previous works have investigated how to model users’ historical behaviors to predict the click-through rate of micro-videos, they are generally based on positive feedback only but overlook the negative which can help understand user preference at a finer granularity. The positive and negative feedback jointly imply the user’s different sentiments on different aspects, where each aspect is one component of a micro-video such as video_scene and video_subject. To this end, we propose an a spect-level s entiment cap sule network(ASCap) for micro-video click-through rate prediction by aggregating both positive and negative feedback, with an attempt to make the prediction more explainable. More specifically, an aspect-specific gating mechanism is firstly utilized to extract the aspect-level features from the target micro-video and the user’s positive and negative feedback. Then, in the following sentiment capsule network, the aspect-level features of the target micro-video are paired with those of positive and negative feedback respectively to identify their sentiments and form the sentiment capsules. Finally, the prediction layer is employed to calculate the overall click probability based on the sentiment capsules. Experimental results on two real-world micro-video datasets demonstrate that the proposed method significantly outperforms the state-of-the-art methods.

Novel coronavirus disease 2019 (COVID-19) is an emerging, rapidly evolving crisis, and the ability to predict prognosis for individual COVID-19 patient is important for guiding treatment. Laboratory examinations were repeatedly measured during hospitalization for COVID-19 patients, which provide the possibility for the individualized early prediction of prognosis. However, previous studies mainly focused on risk prediction based on laboratory measurements at one time point, ignoring disease progression and changes of biomarkers over time. By using historical regression trees (HTREEs), a novel machine learning method, and joint modeling technique, we modeled the longitudinal trajectories of laboratory biomarkers and made dynamically predictions on individual prognosis for 1997 COVID-19 patients. In the discovery phase, based on 358 COVID-19 patients admitted between 10 January and 18 February 2020 from Tongji Hospital, HTREE model identified a set of important variables including 14 prognostic biomarkers. With the trajectories of those biomarkers through 5-day, 10-day and 15-day, the joint model had a good performance in discriminating the survived and deceased COVID-19 patients (mean AUCs of 88.81, 84.81 and 85.62% for the discovery set). The predictive model was successfully validated in two independent datasets (mean AUCs of 87.61, 87.55 and 87.03% for validation the first dataset including 112 patients, 94.97, 95.78 and 94.63% for the second validation dataset including 1527 patients, respectively). In conclusion, our study identified important biomarkers associated with the prognosis of COVID-19 patients, characterized the time-to-event process and obtained dynamic predictions at the individual level.

Background Novel coronavirus disease (COVID-19) is an emerging, rapidly evolving situation. At present, the prognosis of severe and critically ill patients has become an important focus of attention. We strived to develop a prognostic prediction model for severe and critically ill COVID-19 patients.MethodsTo assess the factors associated with the prognosis of those patients, we retrospectively investigated the clinical, laboratory characteristics of confirmed 112 cases of COVID-19 admitted between 21 January to 6 March 2020 from Huangshi Central Hospital, Huangshi Hospital of Traditional Chinese Medicine, and Daye People’s Hospital. We applied machine learning method (survival random forest) to select predictors for 28-day survival and taken into account the dynamic trajectory of laboratory indicators. Results Fifteen candidate prognostic features, including 11 baseline measures (including platelet count (PLT), urea, creatine kinase (CK), fibrinogen, creatine kinase isoenzyme activity, aspartate aminotransferase (AST), activation of partial thromboplastin time (APTT), albumin, standard deviation of erythrocyte distribution width (RBC-SD), neutrophils (%) and red blood cell count (RBC)) and 4 trajectory clusters (changes during hospitalization in the white blood cell (WBC), PLT large cell ratio (P-LCR), PLT distribution width (PDW) and AST), combined with covariates achieved 100% (95%CI: 99%-100%) AUC and reached 87% (95%CI: 84%-91%) AUC in an external validation set. Conclusions Taking advantage of random forest technique and laboratory dynamic measures, we developed a forest model to predict survival outcome of COVID-19 patients, which achieved 87% AUC in the external validation set. Our online tool will help to facilitate the early recognition of patients with high risk.

We describe how to use several machine learning techniques organized in a learning pipeline to segment and identify subcellular structures from cryo electron tomograms. These tomograms are difficult to analyze with traditional segmentation tools. The learning pipeline in our approach starts from supervised learning via a special convolutional neural network trained with simulated data. It continues with semi-supervised reinforcement learning and/or a region merging techniques that try to piece together disconnected components that should belong to the same subcellular structure. A parametric or non-parametric fitting procedure is then used to enhance the segmentation results and quantify uncertainties in the fitting. Domain knowledge is used in generating the training data for the neural network and in guiding the fitting procedure through the use of appropriately chosen priors and constraints. We demonstrate that the approach proposed here work well for extracting membrane surfaces of protein reconstituted liposomes in a cellular environment that contains other artifacts.

In the process of treating pro-diseases with acupuncture, traditional Chinese medicine (TCM) doctors may fine-tune acupuncture prescriptions according to different prior experiences. Different prescriptions will affect the efficiency and effect of acupuncture treatment, and even excessive acupoint selection may cause psychological pressure on patients. We still lack an effective means to analyze the meridian system and acupoint specificity to clarify the mapping relationship between acupoints and diseases. Given the inability of modern medical technology to provide effective evidence support for meridians and acupoints, we combined acupuncture theory with network science for an interdisciplinary discussion. In this paper, we constructed a weighted undirected acupoint-disease network (ADN) based on clinical acupuncture prescription literature and proposed a high-specificity key node mining method based on ADN. Combined with the principle of acupoint selection in TCM, the proposed method balanced the contribution of local areas to the network based on the distribution characteristics of meridians and selected 30 key acupoints with high influence on the global topology according to the evaluation index of key nodes. Finally, we compared the proposed method with the other six classical node importance evaluation algorithms in terms of resolution, network loss, and accuracy. The comprehensive results show that the marked key acupoint nodes make outstanding contributions to the connectivity, topological structure, and weighted benefits of the network, and the stability and specificity of the algorithm guarantee the reliability of the key acupoint nodes. We consider that these key acupoints with high centrality in ADN can be used as core acupoints to help researchers explore targeted and high-impact acupoint combinations under resource constraints and optimize existing acupuncture prescriptions.

With the extensive application of ontology in the fields of information retrieval and artificial intelligence, the ontology-based conceptual similarity calculation becomes a hot topic in ontology research. The essence of ontology learning is to obtain the ontology function through the learning of ontology samples, so as to map the vertices in each ontology graph into real numbers, and finally determine the similarity between corresponding concepts by the difference between real numbers. The essence of ontology mapping is to calculate concepts from different ontologies. In this paper, we introduce new ontology similarity computing in view of stochastic primal dual coordinate method, and two experiments show the effectiveness of our proposed ontology algorithm.

In this paper we study multi-label learning with weakly labeled data, i.e., labels of training examples are incomplete, which commonly occurs in real applications, e.g., image classification, document categorization. This setting includes, e.g., (i) semi-supervised multi-label learning where completely labeled examples are partially known; (ii) weak label learning where relevant labels of examples are partially known; (iii) extended weak label learning where relevant and irrelevant labels of examples are partially known. Previous studies often expect that the learning method with the use of weakly labeled data will improve the performance, as more data are employed. This, however, is not always the cases in reality, i.e., weakly labeled data may sometimes degenerate the learning performance. It is desirable to learn safe multi-label prediction that will not hurt performance when weakly labeled data is involved in the learning procedure. In this work we optimize multi-label evaluation metrics ( $$\hbox {F}_1$$ score and Top-k precision) given that the ground-truth label assignment is realized by a convex combination of base multi-label learners. To cope with the infinite number of possible ground-truth label assignments, cutting-plane strategy is adopted to iteratively generate the most helpful label assignments. The whole optimization is cast as a series of simple linear programs in an efficient manner. Extensive experiments on three weakly labeled learning tasks, namely, (i) semi-supervised multi-label learning; (ii) weak label learning and (iii) extended weak label learning, clearly show that our proposal improves the safeness of using weakly labeled data compared with many state-of-the-art methods.

In biology field, the ontology application relates to a large amount of genetic information and chemical information of molecular structure, which makes knowledge of ontology concepts convey much information. Therefore, in mathematical notation, the dimension of vector which corresponds to the ontology concept is often very large, and thus improves the higher requirements of ontology algorithm. Under this background, we consider the designing of ontology sparse vector algorithm and application in biology. In this paper, using knowledge of marginal likelihood and marginal distribution, the optimized strategy of marginal based ontology sparse vector learning algorithm is presented. Finally, the new algorithm is applied to gene ontology and plant ontology to verify its efficiency.

AUC is an important performance measure that has been used in diverse tasks, such as class-imbalanced learning, cost-sensitive learning, learning to rank, etc. In this work, we focus on one-pass AUC optimization that requires going through training data only once without having to store the entire training dataset. Conventional online learning algorithms cannot be applied directly to one-pass AUC optimization because AUC is measured by a sum of losses defined over pairs of instances from different classes. We develop a regression-based algorithm which only needs to maintain the first and second-order statistics of training data in memory, resulting in a storage requirement independent of the number of training data. To efficiently handle high-dimensional data, we develop two deterministic algorithms that approximate the covariance matrices. We verify, both theoretically and empirically, the effectiveness of the proposed algorithms.

© 2019 Elsevier Ltd This paper presents a machine learning aided stochastic free vibration analysis for functionally graded (FG) bar-type structures through finite element method (FEM). The considered system uncertainties including the constituent material properties, the dimensions of structural members, and the degree of the gradation of the FGM are incorporated. A novel kernel-based machine learning technique, namely the extended support vector regression (X-SVR), is presented to estimate the governing relationship between the uncertain system parameters and the structural natural frequencies. Subsequently, by applying the Monte-Carlo Simulation (MCS) through the established regression model, various types of statistical characteristics (i.e., mean, standard deviation, probability density function or PDF, and cumulative distribution function or CDF) of structural natural frequencies can be effectively established. Four numerical examples including test functions and practically stimulated engineering structures are thoroughly investigated herein to demonstrate the accuracy, applicability, and computational efficiency of the proposed approach.

Having a reliable approximation of heating load (HL) and cooling load (CL) is a substantial task for evaluating the energy performance of buildings (EPB). Also, the appearance of soft computing techniques has made many traditional methods antiquated. Thus, the main effort of this study was to evaluate the capability of several learning methods for appraising the HL and CL of a residential building. To this end, a proper dataset consisting of eight influential factors was provided. To simplify the problem, we executed feature validity by using a correlation-based feature subset selection (CfsSubsetEval) technique. The results of this process showed that wall area, overall height, orientation and glazing area have the most significant impact on the HL and CL simulation. After preparing the suitable dataset, sixteen learning methods namely, elastic net (EN), Gaussian process regression (GPR), least median of squares regression (LMSR), multiple linear regression (MLR), multi-layer perceptron regression (MPR), multi-layer perceptron (MLP), radial basis function regression (RBFR), sequential minimal optimization regression (SMOR), functions XNV, lazy K-star, lazy LWL, rules decision table (RDT), M5Rules, alternating model tree (AMT), directional path consistency (DPC), and Random Forest (RF) were developed in Weka environment to forecast the HL and CL variables. Referring to the results, it was concluded that RF, lazy K-star, RDT and AMT outperform other predictive models. Also, comparing the results with the results of the previous studies showed that the applied feature reduction not only did not disturb the learning process but also has enhanced the performance of models. Also, due to the excellent accuracy of the MLP, a formula was derived from the optimized structure of it to predict the HL and CL variables.

Deep neural networks have witnessed great successes in various real applications, but it requires a large number of labeled data for training. In this paper, we propose tri-net, a deep neural network which is able to use massive unlabeled data to help learning with limited labeled data. We consider model initialization, diversity augmentation and pseudo-label editing simultaneously. In our work, we utilize output smearing to initialize modules, use fine-tuning on labeled data to augment diversity and eliminate unstable pseudo-labels to alleviate the influence of suspicious pseudo-labeled data. Experiments show that our method achieves the best performance in comparison with state-of-the-art semi-supervised deep learning methods. In particular, it achieves 8.30% error rate on CIFAR-10 by using only 4000 labeled examples.

The functional connectome has gained increased attention in the neuroscience community. In general, most network connectivity models are based on correlations between discrete-time series signals that only connect two different brain regions. However, these bivariate region-to-region models do not involve three or more brain regions that form a subnetwork. Here we propose a learning-based method to explore subnetwork biomarkers that are significantly distinguishable between two clinical cohorts. Learning on hypergraph is employed in our work. Specifically, we construct a hypergraph by exhaustively inspecting all possible subnetworks for all subjects, where each hyperedge connects a group of subjects demonstrating highly correlated functional connectivity behavior throughout the underlying subnetwork. The objective function of hypergraph learning is to jointly optimize the weights for all hyperedges which make the separation of two groups by the learned data representation be in the best consensus with the observed clinical labels. We deploy our method to find high order childhood autism biomarkers from rs-fMRI images. Promising results have been obtained from comprehensive evaluation on the discriminative power and generality in diagnosis of Autism.

Automatic rumor detection is technically very challenging. In this work, we try to learn discriminative features from tweets content by following their non-sequential propagation structure and generate more powerful representations for identifying different type of rumors. We propose two recursive neural models based on a bottom-up and a top-down tree-structured neural networks for rumor representation learning and classification, which naturally conform to the propagation layout of tweets. Results on two public Twitter datasets demonstrate that our recursive neural models 1) achieve much better performance than state-of-the-art approaches; 2) demonstrate superior capacity on detecting rumors at very early stage.

In this paper, a joint machine learning and game theory modeling (MLGT) framework is proposed for inter frame coding tree unit (CTU) level bit allocation and rate control (RC) optimization in high efficiency video coding (HEVC). First, a support vector machine-based multi-classification scheme is proposed to improve the prediction accuracy of CTU-level rate-distortion (R-D) model. The legacy “chicken-and-egg” dilemma in video coding is proposed to be overcome by the learning-based R-D model. Second, a mixed R-D model-based cooperative bargaining game theory is proposed for bit allocation optimization, where the convexity of the mixed R-D model-based utility function is proved, and Nash bargaining solution is achieved by the proposed iterative solution search method. The minimum utility is adjusted by the reference coding distortion and frame-level quantization parameter (QP) change. Finally, intra frame QP and inter frame adaptive bit ratios are adjusted to make inter frames have more bit resources to maintain smooth quality and bit consumption in the bargaining game optimization. Experimental results demonstrate that the proposed MLGT-based RC method can achieve much better R-D performances, quality smoothness, bit rate accuracy, buffer control results, and subjective visual quality than the other state-of-the-art one-pass RC methods, and the achieved R-D performances are very close to the performance limits from the FixedQP method.

In this article, we present a sparse vector learning algorithm for ontology similarity measure and ontology mapping by virtue of accelerated first-order technology. The main procedure of our iterative algorithm is relying on prox- imity operator computation, Picard-Opial process and accelerated first-order tricks. The simulation experimental results show that the new proposed algorithm has high efficiency and accuracy in ontology similarity measure and ontology mapping in plant science and university application.

Analyzing geological drilling hole images acquired by Axial View Panoramic Borehole Televiewer (APBT) is a key step to explore the geological structure in a geological exploration. Conventionally, the borehole images are examined by technicians, which is inefficient and subjective. In this paper, three dominant types of borehole-wall images on coal-rock mass structure, namely, border images, fracture images and intact rock mass images are mainly studied. The traditional image classification methods based on unified feature extraction algorithm and single classifier is not effect for the borehole images. Therefore, this paper proposes a novel two-stage classification approach to improve the classification performance of borehole images. In the first-stage classification, the border images are identified from three kinds of images based on texture features and gray-scale histograms features. For the remaining two types of images, in the second-stage classification, Gabor filter is first applied to segment the region of interest (ROI) (such as microfracture, absciss layer and horizontal cracks, etc.) and the central interference region. Then, using the same feature vector after eliminating the central interference region, fracture images are separated from intact rock mass images. We test our two-stage classification system with real borehole images. The results of experimental show that the two-stage classification method can effectively classify three major borehole-wall images with the correction rate of 95.55% in the first stage and 95% in the second stage.

How fake news goes viral via social media? How does its propagation pattern differ from real stories? In this paper, we attempt to address the problem of identifying rumors, i.e., fake information, out of microblog posts based on their propagation structure. We firstly model microblog posts diffusion with propagation trees, which provide valuable clues on how an original message is transmitted and developed over time. We then propose a kernel-based method called Propagation Tree Kernel, which captures high-order patterns differentiating different types of rumors by evaluating the similarities between their propagation tree structures. Experimental results on two real-world datasets demonstrate that the proposed kernel-based approach can detect rumors more quickly and accurately than state-of-the-art rumor detection models.

Functional magnetic resonance imaging (fMRI) provides a non-invasive way to investigate brain activity. Recently, convergent evidence shows that the correlations of spontaneous fluctuations between two distinct brain regions dynamically change even in resting state, due to the condition-dependent nature of brain activity. Thus, quantifying the patterns of functional connectivity (FC) in a short time period and changes of FC over time can potentially provide valuable insight into both individual-based diagnosis and group comparison. In light of this, we propose a novel computational method to robustly estimate both static and dynamic spatial-temporal connectivity patterns from the observed noisy signals of individual subject. We achieve this goal in two folds: (1) Construct static functional connectivity across brain regions. Due to low signal-to-noise ratio induced by possible non-neural noise, the estimated FC strength is very sensitive and it is hard to define a good threshold to distinguish between real and spurious connections. To alleviate this issue, we propose to optimize FC which is in consensus with not only the low level region-to-region signal correlations but also the similarity of high level principal connection patterns learned from the estimated link-to-link connections. Since brain network is intrinsically sparse, we also encourage sparsity during FC optimization. (2) Characterize dynamic functional connectivity along time. It is hard to synchronize the estimated dynamic FC patterns and the real cognitive state changes, even using learning-based methods. To address these limitations, we further extend above FC optimization method into the spatial-temporal domain by arranging the FC estimations along a set of overlapped sliding windows into a tensor structure as the window slides. Then we employ low rank constraint in the temporal domain assuming there are likely a small number of discrete states that the brain transverses during a short period of time. We applied the learned spatial-temporal patterns from fMRI images to identify autism subjects. Promising classification results have been achieved, suggesting high discrimination power and great potentials in computer assisted diagnosis.

In recent years there has been a growing interest in text quantification, a supervised learning task where the goal is to accurately estimate, in an unlabelled set of items, the prevalence (or "relative frequency") of each class c in a predefined set C. Text quantification has several applications, and is a dominant concern in fields such as market research, the social sciences, political science, and epidemiology. In this paper we tackle, for the first time, the problem of ordinal text quantification, defined as the task of performing text quantification when a total order is defined on the set of classes; estimating the prevalence of "five stars" reviews in a set of reviews of a given product, and monitoring this prevalence across time, is an example application. We present OQT, a novel tree-based OQ algorithm, and discuss experimental results obtained on a dataset of tweets classified according to sentiment strength.

Great successes of deep neural networks have been witnessed in various real applications. Many algorithmic and implementation techniques have been developed, however, theoretical understanding of many aspects of deep neural networks is far from clear. A particular interesting issue is the usefulness of dropout, which was motivated from the intuition of preventing complex co-adaptation of feature detectors. In this paper, we study the Rademacher complexity of different types of dropout, and our theoretical results disclose that for shallow neural networks (with one or none hidden layer) dropout is able to reduce the Rademacher complexity in polynomial, whereas for deep neural networks it can amazingly lead to an exponential reduction of the Rademacher complexity., Comment: 20 pagea

Sentiment classification has become a ubiquitous enabling technology in the Twittersphere, since classifying tweets according to the sentiment they convey towards a given entity (be it a product, a person, a political party, or a policy) has many applications in political science, social science, market research, and many others. In this paper, we contend that most previous studies dealing with tweet sentiment classification (TSC) use a suboptimal approach. The reason is that the final goal of most such studies is not estimating the class label (e.g., Positive, Negative, or Neutral) of individual tweets, but estimating the relative frequency (a.k.a. "prevalence") of the different classes in the dataset. The latter task is called quantification, and recent research has convincingly shown that it should be tackled as a task of its own, using learning algorithms and evaluation measures different from those used for classification. In this paper, we show (by carrying out experiments using two learners, seven quantification-specific algorithms, and 11 TSC datasets) that using quantification-specific algorithms produces substantially better class frequency estimates than a state-of-the-art classification-oriented algorithm routinely used in TSC. We thus argue that researchers interested in tweet sentiment prevalence should switch to quantification-specific (instead of classification-specific) learning algorithms and evaluation measures.

In view of modeling accuracy problems of General Regression Neural Network (GRNN), the improved GRNN has been used to forecast wind speed. Firstly, K-fold cross validation was used to select smooth parameter of GRNN, and the influence of K value was analyzed. Then, Markov Chain (MC) was introduced to correct the results of GRNN to improve the accuracy of GRNN. And C-average clustering algorithm was used to state division, the number of state and correction results were emphatically discussed. Finally, the improved GRNN was applied to wind speed forecasting by use of the actual data gathered from a wind farm in Shanxi province to further test the proposed method. The experimental results show that the superiority of the proposed method compared with GRNN.

Wind speed forecasting has significant influence on wind energy development. In the paper, a least square support vector machine with a novel optimization algorithm was used to improve the performances of wind speed forecasting model. Coupled simulated annealing (CSA) and simplex algorithm were combined in the novel optimization algorithm to optimize parameters of LSSVM forecasting model. Firstly, parameters were optimized by CSA in global scope. Then, parameters which got from CSA were optimized by simplex algorithm to get the best parameters. Finally, the LSSVM model with best parameters was applied to wind speed forecasting. Based on the data obtained from a wind farm in Shanxi province, the simulation results show that comparing with the support vector machine (SVM) model with grid-search and the LSSVM model with particle swarm optimization, the proposed model has better performances on accuracy and training time, thereby it helps make reasonable decisions for power scheduling and dispatch. Keywords-wind speed forecasting; least squares support vector machine (LSSVM); coupled simulated annealing algorithm; simplex algorithm

Sentiment classification has become a ubiquitous enabling technology in the Twittersphere, since classifying tweets according to the sentiment they convey towards a given entity (be it a product, a person, a political party, or a policy) has many applications in political science, social science, market research, and many others. In this paper we contend that most previous studies dealing with tweet sentiment classification (TSC) use a suboptimal approach. The reason is that the final goal of most such studies is not estimating the class label (e.g., Positive, Negative, or Neutral) of individual tweets, but estimating the relative frequency (a.k.a. "prevalence") of the different classes in the dataset. The latter task is called quantification, and recent research has convincingly shown that it should be tackled as a task of its own, using learning algorithms and evaluation measures different from those used for classification. In this paper we show, on a multiplicity of TSC datasets, that using a quantification-specific algorithm produces substantially better class frequency estimates than a state-of-the-art classification-oriented algorithm routinely used in TSC. We thus argue that researchers interested in tweet sentiment prevalence should switch to quantification-specific (instead of classification-specific) learning algorithms and evaluation measures.