Search

Many practical applications, such as social media and monitoring system, will constantly generate streaming data, which has problems of instability, lack of labels and multiclass imbalance. In order to solve these problems, a cluster-based active learning method is proposed to achieve data stream classification. Firstly, a label query strategy combining marginal threshold matrix is proposed, which selects difficult to classify or potential concept drift samples for marking, to solve the problem of high cost label and unbalanced data. Secondly, dynamic maintenance of a group of micro clusters, by adjusting the weight of micro clusters in the model, correctly reflects the current data distribution, and finally, uses the buffer to store new micro clusters to participate in the update of the model, to adapt to the new data environment. Experimental results on three real data sets and three synthetic data sets show that compared with the classical data stream classification algorithm, it is less affected by concept drift and has higher classification accuracy than the online semi-supervised learning algorithm ADSM. The average accuracy of the six datasets increased by 5.56%, 2.32%, 1.77%, 1.83%, 3.78%, and 2.04%, respectively. The model processes data streams online and improves classification performance with less memory consumption. [ABSTRACT FROM AUTHOR]

Internet of Things (IoT) realizes the interconnection of heterogeneous devices by the technology of wireless and mobile communication. The data of target regions are collected by widely distributed sensing devices and transmitted to the processing center for aggregation and analysis as the basis of IoT. The quality of IoT services usually depends on the accuracy and integrity of data. However, due to the adverse environment or device defects, the collected data will be anomalous. Therefore, the effective method of anomaly detection is the crucial issue for guaranteeing service quality. Deep learning is one of the most concerned technology in recent years which realizes automatic feature extraction from raw data. In this article, the integrated model of the convolutional neural network (CNN) and recurrent autoencoder is proposed for anomaly detection. Simple combination of CNN and autoencoder cannot improve classification performance, especially, for time series. Therefore, we utilize the two-stage sliding window in data preprocessing to learn better representations. Based on the characteristics of the Yahoo Webscope S5 dataset, raw time series with anomalous points are extended to fixed-length sequences with normal or anomaly label via the first-stage sliding window. Then, each sequence is transformed into continuous time-dependent subsequences by another smaller sliding window. The preprocessing of the two-stage sliding window can be considered as low-level temporal feature extraction, and we empirically prove that the preprocessing of the two-stage sliding window will be useful for high-level feature extraction in the integrated model. After data preprocessing, spatial and temporal features are extracted in CNN and recurrent autoencoder for the classification in fully connected networks. Empiric results show that the proposed model has better performances on multiple classification metrics and achieves preferable effect on anomaly detection. [ABSTRACT FROM AUTHOR]

People's increasingly frequent online activity has generated a large number of reviews, whereas fake reviews can mislead users and harm their personal interests. In addition, it is not feasible to label reviews on a large scale because of the high cost of manual labeling. Therefore, to improve the detection performance by utilizing the unlabeled reviews, this article proposes a fake reviews detection model based on vertical ensemble tri-training and active learning (VETT-AL). The model combines the features of review text with the user behavior features as feature extraction. In the VETT-AL algorithm, the iterative process is divided into two parts: vertical integration within the group and horizontal integration among the groups. The intra-group integration is to integrate three original classifiers by using the previous iterative models of the classifiers. The inter-group integration is to adopt the active learning based on entropy to select the data with the highest confidence and label it, and as the result of that, the second generation classifiers are trained by the traditional process to improve the accuracy of the label. Experimental results show that the proposed model has a good classification performance. [ABSTRACT FROM AUTHOR]

In order to receive efficient personalized recommendation, users have to provide personal information to service providers. However, in this process, personal private data are in an extremely dangerous situation. Personalized recommendation technology based on privacy protection can enable users to enjoy personalized recommendations, while private data are also protected. In this paper, an efficient privacy-preserving collaborative filtering algorithm is proposed, which is based on differential privacy protection and time factor. The proposed method used the MovieLens data set in the experiment. Experimental results showed that the proposed method can effectively protect the private data, but the accuracy of recommendation is slightly inferior than the traditional collaborative filtering algorithm. [ABSTRACT FROM AUTHOR]

Location-based recommendation services can provide users with convenient services, but this requires monitoring and collecting a large amount of location information. In order to prevent location information from being leaked after monitoring and collection, location privacy must be effectively protected. Therefore, this paper proposes a privacy protection method based on location sensitivity for location recommendation. This method uses location trajectories and check-in frequencies to set a threshold so as to classify location sensitivity levels. The corresponding privacy budget is then assigned based on the sensitivity to add Laplace noise that satisfies the differential privacy. Experimental results show that this method can effectively protect the user's location privacy and reduce the impact of differential privacy noise on service quality. [ABSTRACT FROM AUTHOR]

Finite state automata are widely used in firewalls, data detection and content audit systems to match complex sets of regular expressions in network packets. However, with the continuous increase in the types of network contents and network traffics in recent years, the deep packet inspection systems based on finite state automata also require regular engines for less memory consumption and higher operating speed. This paper analyzes the feature and problem of finite state automata and improves non-deterministic finite automata by reducing the conversion edge to reduce the memory usage. The experiment results which are made by real-world dataset show that the memory usage is reduced more than half. [ABSTRACT FROM AUTHOR]

Intrusion detection provides important protection for network security and anomaly detection as a type of intrusion detection, which can recognize the pattern of normal behaviors and label the behaviors which departure from normal pattern as anomaly behaviors. The updating of network equipment and broadband speed makes the data mining object change from static data sets to dynamic data streams. We think that the traditional methods based on data set do not satisfy the needs of dynamic network environment. The network data stream is temporal and cannot be treated as static data set. The concept and distribution of data objects is variety in different time stamps and the changing is unpredictable. Therefore, we propose an improved data stream clustering algorithm and design the anomaly detection model according to the improved algorithm. The established model can be modified with the changing of data stream and detect anomaly behaviors in time. [ABSTRACT FROM AUTHOR]

With the continuous development of network technology, the attack has become the biggest threat to the stable operation of the network. Intrusion detection technology is a proactive safety protection measure which provides real-time monitoring of internal attacks, external attacks, and misuse. Traditional intrusion detection system is short of adaptability due to the complication and scale of the network. The main problem is that the real-time performance of the network is poor and the reliability is not high. This paper designs the intrusion detection mechanism combined with data stream clustering algorithm and intrusion detection system to solve the problem in processing a large amount of high-speed data streams. The performance of processing data streams is improved through the clustering algorithm based on density and the sliding window and the experiments show that the intrusion detection efficiency is higher than DenStream algorithm. [ABSTRACT FROM AUTHOR]

Personal recommendation technology is becoming a useful and popular solution to solve the problem of information overload with the popularity of big data cloud services. But most recommendation algorithms pay too much attention to the similarity to focus on the social trust between users. So this paper focus on the research of hybrid Recommendation algorithm for big data based on the optimization combining with the similarity and trust in sociology. In this paper, we introduced some user trust models including trust path model and loop trust model, and then we took these models into the calculation of mixed weighting. The experiment results show that the recommendation algorithm considering the trust models has the higher accuracy than the traditional recommendation algorithm, and we have a 2% increase in both MEA (Mean Absolute Error) and RMSE (Root Mean Square Error). [ABSTRACT FROM AUTHOR]

Depending on the use of the Internet and network, data-stream classification has been applied in the intrusion detection field. Due to unlimited and difficult storage features, the routine classification algorithm (eg. C4.5, currently widely used classification algorithm with higher classification accuracy) tends to incorrect classification and memory leaks. In this paper, we propose an improved Hoeffding tree data-stream classification algorithm, Hoeffding-ID and apply it to the network data-stream process of the intrusion detection field. Experimental results shows that the Hoeffding-ID algorithm has relative high detection accuracy, low positives rate and memory usage not increasing with the data samples. [ABSTRACT FROM AUTHOR]

Intrusion detection is a kind of security mechanism which is used to detect attacks and intrusion behaviors. Due to the low accuracy and the high false positive rate of the existing clonal selection algorithms applied to intrusion detection, in this paper, we proposed a feature selection method for improved clonal algorithm. The improved method detects the intrusion behavior by selecting the best individual overall and clones them. Experimental results show that the feature selection algorithm is better than the traditional feature selection algorithm on the different classifiers, and it is shown that the final detection results are better than traditional clonal algorithm with 99.6% accuracy and 0.1% false positive rate. [ABSTRACT FROM AUTHOR]

We reported complete mitochondrial genome of neotropical Parnell's mustached batPteronotus parnellii(16,564 bp in length). Its mitochondrial genome consisted of 13 protein-coding genes, 22 transfer RNA genes, 2 ribosomal RNA genes and 1 control region. The base composition of mtDNA was A (33.1%), G (13.7%), C (27.3%), T (25.9%), and the percentage of A and T (59.0%) was higher than G and C. All the genes inP. parnelliiwere distributed on the H-strand, except for theND6subunit gene and 10 tRNA genes, which were encoded on the L-strand. The new sequenced mitochondrial genome will contribute to our understanding of the systematic evolution in the bats. [ABSTRACT FROM PUBLISHER]

Anomaly detection, which is a method of intrusion detection, detects anomaly behaviors and protects network security. Data mining technology has been integrated to improve the performance of anomaly detection and some algorithms have been improved for anomaly detection field. We think that most data mining algorithms are analyzed on static data sets and ignore the influence of dynamic data streams. Data stream is the potentially unbounded, ordered sequence of data objects which arrive over time. The entire data objects cannot be stored and they need to be handled in one-time scanning. The data distribution of data stream may change over time and this phenomenon is called concept drift. The properties of data stream make analysis method different from the method based on data set and the analysis model is required to be updated immediately when concept drift occurs. In this paper, we summarize the characteristics of data stream, compare the difference between data stream and data set, discuss the problems of data stream mining and propose some corresponding strategies. [ABSTRACT FROM AUTHOR]

The continuous development of Internet technology makes the network intrusion detection technology get more and more attention. Deep packet inspection technology as an effective network intrusion detection technology can play a huge role in network security. Deep packet inspection technology is a kind of network intrusion detection technology applied to the application layer in detail, rather than only detecting the port information of the packet. The regular expression matching technology is a key technology in deep packet inspection because of the rich semantics and flexibility of regular expressions. However, a huge number of transfer edges exist when the matching algorithm is being applied, which will lead to an increase in memory usage of the algorithm. In this paper, we propose an improved method of concatenating transfer edges. By using character interval, several consecutive characters are represented by character intervals, which can reduce the number of transfer edges effectively. In addition, a comparison experiment is given to compare the two methods which are before and after the improvement. It shows that the number of transfer edges can be reduced to 10% of that before improvement and the efficiency of deep packet inspection is improved. [ABSTRACT FROM AUTHOR]

The rising use of online media has changed the social customs of the public. Users have become accustomed to sharing daily experiences and publishing personal opinions on social networks. Social data carrying emotion and attitude has provided significant decision support for numerous tasks in sentiment analysis. Conventional methods for sentiment classification only concern textual modality and are vulnerable to the multimodal scenario, while common multimodal approaches only focus on the interactive relationship among modalities without considering unique intra-modal information. A hybrid fusion network is proposed in this paper to capture both inter-modal and intra-modal features. Firstly, in the stage of representation fusion, a multi-head visual attention is proposed to extract accurate semantic and sentimental information from textual contents, with the guidance of visual features. Then, multiple base classifiers are trained to learn independent and diverse discriminative information from different modal representations in the stage of decision fusion. The final decision is determined based on fusing the decision supports from base classifiers via a decision fusion method. To improve the generalization of our hybrid fusion network, a similarity loss is employed to inject decision diversity into the whole model. Empiric results on five multimodal datasets have demonstrated that the proposed model achieves higher accuracy and better generalization capacity for multimodal sentiment analysis. [ABSTRACT FROM AUTHOR]

A model predictive control (MPC) technique is proposed for the frequency stabilization of transverse Zeeman He–Ne laser. The beat frequency locking method is applied as the criterion of frequency stabilization. When heated by the voltage output of the feedback electronic circuit, the copper wire twisted around the laser tube is applied to maintain the length of the laser cavity and stabilize the frequency of the laser. Four steps for MPC design, i.e., model setup, output prediction, selection of reference path, and calculation of control quality, have been introduced. By this means, a single-chip microprocessor (80C196) generates a pulse width modulation wave that is transferred into a heating signal. The calibration result shows that the frequency stability of our proposed transverse Zeeman laser reaches 5.5×10-11 by means of the MPC method. [ABSTRACT FROM AUTHOR]

Nondiffracting beam is suitable for large-scale straightness measurement because it is much less influenced by atmospheric turbulence than other beams. Based on the energy conservation theory, a binary optical element (BOE) that is used to transform an incident plane wave or divergent spherical wave into a nondiffracting beam was fabricated. We compared intensity distribution of the nondiffracting beam at different propagation distances between a plane wave incidence and a divergent spherical wave incidence based on the scalar diffraction theory and experiment results, and found that by using the same BOE, the propagation range of nondiffracting beam generated by a divergent spherical wave is longer than that generated by a plane wave. [Copyright &y& Elsevier]

A multi-exposure of color fringes method has been developed to improve the capture speed of a conventional color CCD camera. In the method, four groups of projected fringe patterns encoded with different colors and different directions are stored in one CCD frame. Therefore the capture frequency of the conventional CCD can be improved to 200 Hz. It is available to measure the insect wings with low beating frequency, such as dragonfly, moth, or butterfly, whose beating frequency is about 30–40 Hz. We have used the method to measure the beating motion of a moth successfully. [Copyright &y& Elsevier]

A video tracking system based on fringe tracking and camera tracking is developed to simultaneously measure the spatial position and the body deformation of a swimming fish. The control module of the system consists of a fringe pattern tracking system including one tracking mirror and a fringe pattern projector, and a video tracking system including another tracking mirror and a high-speed camera. In the control module, a target trajectory prediction algorithm is used to predict the position of a swimming fish. The two mirrors mounted on two step motors are used to track a swimming fish. The rotation angles of the tracking mirrors are recorded simultaneously. Body position and deformation are calculated from the distorted fringes based on triangulation. The relative accuracy on the body deformation is below 1%. We successfully apply the system to a swimming fish. [ABSTRACT FROM AUTHOR]

We introduce a simple ray tracing method for measuring the vertex radius of curvature in situ, based on projecting spot array rays onto the surface of an aspheric mirror, and receiving the incident and the reflected rays. The method was used to measure the vertex radiuses of curvature of a spherical mirror and a paraboloidal mirror, the relative accuracy of better than 0.5% has been obtained. [Copyright &y& Elsevier]

A novel roll angle interferometer using a transverse Zeeman laser (TZL) is proposed in this paper. The orthogonally linearly-polarized light from TZL is changed to be slightly elliptically-polarized light by a quarter wave plate (QWP) when the principal axis of the QWP is at a small angle with the basis of the direction of polarization. The beam then passes through a half wave plate (HWP) that is employed as a sensor. After that it is reflected by a right angle prism (RAP) and returns to the HWP again. Owing to the reasonable transformation of coordinates by the prism, the change of direction of polarization is accumulated twice when the beam passes forward and back through the HWP. Theoretical analysis and experimental results show that the sensitivity of roll measurement can reach sub-second and be improved four times than that we had reported. This system of implementing roll measurement takes the advantage of compactness, stable performance, low cost and robust to the environment disturbance. [Copyright &y& Elsevier]

The heterodyne interferometric system based on phase measurement plays a significant role in modern metrology fields owing to its convenience of signal processing. In this paper, a phase measurement scheme to determine the fraction of the interference fringe modulated by a PZT in a differential single-frequency interferometer is investigated. To overcome the problem of phase jump and inhomogeneous period of the signal, a differential phase discrimination approach with an external trigger signal is presented. The phase measurement system is carried out using a field-programmable gate array (FPGA) device. It is testified by experiments that, over a frequency range of 300 Hz-10 kHz, the phase measurement system has a stability of ±0.015° and a maximum error of ± 0.07°. [ABSTRACT FROM AUTHOR]

The camber deformations of the flapping insect wing provide important information for lift force analysis. Recent measurements are nearly all based on the hypothesis of the rigid wing, neglecting the camber deformation. In this article, a measurement system based on fringe pattern projection is introduced. Projected comb fringes, whose intensity function is near a comb function, were used instead of conventional sine fringes because of their high intensity and sharpness, allowing us to take the distorted fringes on the transparent dragonfly wing easily. The camber deformation and torsion angle of a flapping dragonfly wing can be measured without hypothesizing that the wing is a rigid plate. The experimental result shows that the camber deformation is different during upstroke and downstroke. It is very important for the explanation of the lift force. © 2001 American Institute of Physics. [ABSTRACT FROM AUTHOR]