Recently, the Static Heterogeneous Particle Swarm Optimization (SHPSO) has been studied by more and more researchers. In SHPSO, the different search behaviours assigned to particles during initialization do not change during the search process. As a consequence of this, the inappropriate population size of exploratory particles could leave the SHPSO with great difficulties of escaping local optima. This motivated our attempt to improve the performance of SHPSO by introducing the dynamic heterogeneity. The self-adaptive heterogeneity is able to alter its heterogeneous structure according to some events caused by the behaviour of the swarm. The proposed triggering events are confirmed by keeping track of the frequency of the unchanged global best position (p_g) for a number of iterations. This information is then used to select a new heterogeneous structure when p_g is considered stagnant. According to the different types of heterogeneity, DHPSO-d and DHPSO-p are proposed in this paper. In, particles dynamically use different rules for updating their position when the triggering events are confirmed. In DHPSO-p, a global gbest model and a pairwise connection model are automatically selected by the triggering configuration. In order to investigate the scalability of and DHPSO-p, a series of experiments with four state-of-the-art algorithms are performed on ten well-known optimization problems. The scalability analysis of and DHPSO-p reveals that the dynamic self-adaptive heterogeneous structure is able to address the exploration-exploitation trade-off problem in PSO, and provide the excellent optimal solution of a problem simultaneously.

In this paper, we present an FPGA hardware implementation for a phylogenetic tree reconstruction with a maximum parsimony algorithm. We base our approach on a particular stochastic local search algorithm that uses the Progressive Neighborhood and the Indirect Calculation of Tree Lengths method. This method is widely used for the acceleration of the phylogenetic tree reconstruction algorithm in software. In our implementation, we define a tree structure and accelerate the search by parallel and pipeline processing. We show results for eight real-world biological datasets. We compare execution times against our previous hardware approach, and TNT, the fastest available parsimony program, which is also accelerated by the Indirect Calculation of Tree Lengths method. Acceleration rates between 34 to 45 per rearrangement, and 2 to 6 for the whole search, are obtained against our previous hardware approach. Acceleration rates between 2 to 36 per rearrangement, and 18 to 112 for the whole search, are obtained against TNT.

Class imbalance has drawn much attention of researchers in software defect prediction. In practice, the performance of defect prediction models may be affected by the class imbalance problem. In this paper, we present an approach to evaluating the performance stability of defect prediction models on imbalanced datasets. First, random sampling is applied to convert the original imbalanced dataset into a set of new datasets with different levels of imbalance ratio. Second, typical prediction models are selected to make predictions on these new constructed datasets, and Coefficient of Variation (C·V) is used to evaluate the performance stability of different models. Finally, an empirical study is designed to evaluate the performance stability of six prediction models, which are widely used in software defect prediction. The results show that the performance of C4.5 is unstable on imbalanced datasets, and the performance of Naive Bayes and Random Forest are more stable than other models.

The proliferation of diverse kinds of programming languages and platforms makes it a common need to have the same functionality implemented in different languages for different platforms, such as Java for Android applications and C# for Windows phone applications. Although versions of code written in different languages appear syntactically quite different from each other, they are intended to implement the same software and typically contain many code snippets that implement similar functionalities, which we call cross-language clones. When the version of code in one language evolves according to changing functionality requirements and/or bug fixes, its cross-language clones may also need be changed to maintain consistent implementations for the same functionality. Thus, it is needed to have automated ways to locate and track cross-language clones within the evolving software. In the literature, approaches for detecting cross-language clones are only for languages that share a common intermediate language (such as the .NET language family) because they are built on techniques for detecting single-language clones. To extend the capability of cross-language clone detection to more diverse kinds of languages, we propose a novel automated approach, CLCMiner, without the need of an intermediate language. It mines such clones from revision histories, based on our assumption that revisions to different versions of code implemented in different languages may naturally reflect how programmers change cross-language clones in practice, and that similarities among the revisions (referred to as clones in diffs or diff clones) may indicate actual similar code. We have implemented a prototype and applied it to ten open source projects implementations in both Java and C#. The reported clones that occur in revision histories are of high precisions (89% on average) and recalls (95% on average). Compared with token-based code clone detection tools that can treat code as plain texts, our tool can detect significantly more cross-language clones. All the evaluation results demonstrate the feasibility of revision-history based techniques for detecting cross-language clones without intermediates and point to promising future work.

With the rapid development of information and Web technologies, people are facing ‘information overload’ in their daily lives. The personalized recommendation system (PRS) is an effective tool to assist users extract meaningful information from the big data. Collaborative filtering (CF) is one of the most widely used personalized recommendation techniques to recommend the personalized products for users. However, the conventional CF technique has some limitations, such as the low accuracy of of similarity calculation, cold start problem, etc. In this paper, a PRS model based on the Support Vector Machine (SVM) is proposed. The proposed model not only considers the items' content information, but also the users' demographic and behavior information to fully capture the users' interests and preferences. An improved Particle Swarm Optimization (PSO) algorithm is also proposed to improve the performance of the model. The efficiency of the proposed method is verified by multiple benchmark datasets.

RAID has been widely deployed in disk array storage systems to manage both performance and reliability simultaneously. RAID conducts two performance-critical operations during disk failures known as degraded reads/writes and recovery process. Before the recovery process is complete, reads and writes are degraded because data is reconstructed using data redundancy. The performance of degraded reads/writes is critical in order to meet stipulations in customer service level agreements (SLAs), and the recovery process affects the reliability of a storage system considerably. Both operations require fast data reconstruction. Among the erasure codes for fast reconstruction, Local Reconstruction Codes (LRC) are known to offer the best (or optimal) trade-off between storage overhead, fault tolerance, and the number of disks involved in reconstruction. Originally, LRC was designed for fast reconstruction in distributed cloud storage systems, in which network traffic is a major bottleneck during reconstruction. Thus, LRC focuses on reducing the number of disks involved in data reconstruction, which reduces network traffic. However, we observe that when LRC is applied to primary array storage systems, a major bottleneck in reconstruction results from uneven disk utilization. In other words, underutilized disks can no longer receive I/O requests as a result of the bottleneck of overloaded disks. Uneven disk utilization in LRC is due to its dedicated group partitioning policy to achieve the Maximally Recoverable property. In this paper, we present Distributed Reconstruction Codes (DRC) that support fast reconstruction in primary array storage systems. DRC is designed with group shuffling policy to solve the problem of uneven disk utilization. Experiments on real-world workloads show that DRC using global parity rotation (DRC-G) improves degraded performance by as much as 72% compared to RAID-6 and by as much as 35% compared to LRC under the same reliability. In addition, our study shows that DRC-G reduces the recovery process completion time by as much as 52% compared to LRC.

Mobile traffic is experiencing tremendous growth, and this growing wave is no doubt increasing the use of radio component of mobile devices, resulting in shorter battery lifetime. In this paper, we present an Energy-Aware Download Method (EDM) based on the Markov Decision Process (MDP) to optimize the data download energy for mobile applications. Unlike the previous download schemes in literature that focus on the energy efficiency by simply delaying the download requests, which often leads to a poor user experience, our MDP model learns off-line from a set of training download workloads for different user patterns. The model is then integrated into the mobile application to deal the download request at runtime, taking into account the current battery level, LTE reference signal receiving power (RSRP), reference signal signal to noise radio (RSSNR) and task size as input of the decision process, and maximizes the reward which refers to the expected battery life and user experience. We evaluate how the EDM can be used in the context of a real file downloading application over the LTE network. We obtain, on average, 20.3%, 15% and 45% improvement respectively for energy consumption, latency, and performance of energy-delay trade off, when compared to the Android default download policy (Minimum Delay).

In order to prevent the synonym substitution breaking the balance among frequencies of synonyms and improve the statistical undetectability, this paper proposed a novel linguistic steganography based on synonym run-length encoding. Firstly, taking the relative word frequency into account, the synonyms appeared in the text are digitized into binary values and expressed in the form of runs. Then, message are embedded into the parities of runs' lengths by self-adaptively making a positive or negative synonym transformation on boundary elements of two adjacent runs, while preserving the number of relative high and low frequency synonyms to reduce the embedding distortion. Experimental results have shown that the proposed synonym run-length encoding based linguistic steganographic algorithm makes fewer changes on the statistical characteristics of cover texts than other algorithms, and enhances the capability of anti-steganalysis.

Process variation has become prominent in the advanced CMOS technology, making the timing of fabricated circuits more uncertain. In this paper, we propose a Layout-Aware Path Selection (LAPS) technique to accurately estimate the circuit timing variation from a small set of paths. Three features of paths are considered during the path selection. Experiments conducted on benchmark circuits with process variation simulated with VARIUS show that, by selecting only hundreds of paths, the fitting errors of timing distribution are kept below 5.3% when both spatial correlated and spatial uncorrelated process variations exist.

In text-independent online writer identification, the Gaussian Mixture Model(GMM) writer model trained with the GMM-Universal Background Model(GMM-UBM) framework has acquired excellent performance. However, the system assumes the items in the observation sequence are independent, which neglects the dynamic information between observations. This work shows that although in the text-independent application, the dynamic information between observations is still important for writer identification. In order to extend the GMM-UBM system to use the dynamic information, the hidden Markov model(HMM) with Gaussian observation model is used to model each writer's handwriting in this paper and a new training schematic is proposed. In particular, the observation model parameters of the writer specific HMM are set with the Gaussian component parameters of the GMM writer model trained with the GMM-UBM framework and the state transition matrix parameters are learned from the writer specific data. Experiments show that incorporating the dynamic information is capable of improving the performance of the GMM-based system and the proposed training method is effective for learning the HMM writer model.

Self-paced e-learning provides much more freedom in time and locale than traditional education as well as diversity of learning contents and learning media and tools. However, its limitations must not be ignored. Lack of information on learners' states is a serious issue that can lead to severe problems, such as low learning efficiency, motivation loss, and even dropping out of e-learning. We have designed a novel e-learning support system that can visually observe learners' non-verbal behaviors and estimate their learning states and that can be easily integrated into practical e-learning environments. Three pairs of internal states closely related to learning performance, concentration-distraction, difficulty-ease, and interest-boredom, were selected as targets of recognition. In addition, we investigated the practical problem of estimating the learning states of a new learner whose characteristics are not known in advance. Experimental results show the potential of our system.

Restoration is an important area in improving the visual quality, and lays the foundation for accurate object detection or terrain classification in image analysis. In this paper, we introduce Beta process priors into hierarchical sparse Bayesian learning for recovering underlying degraded hyperspectral images (HSI), including suppressing the various noises and inferring the missing data. The proposed method decomposes the HSI into the weighted summation of the dictionary elements, Gaussian noise term and sparse noise term. With these, the latent information and the noise characteristics of HSI can be well learned and represented. Solved by Gibbs sampler, the underlying dictionary and the noise can be efficiently predicted with no tuning of any parameters. The performance of the proposed method is compared with state-of-the-art ones and validated on two hyperspectral datasets, which are contaminated with the Gaussian noises, impulse noises, stripes and dead pixel lines, or with a large number of data missing uniformly at random. The visual and quantitative results demonstrate the superiority of the proposed method.

Periodic interference frequently affects the measurement of small signals and causes problems in clinical diagnostics. Adaptive filters can be used as potential tools for cancelling such interference. However, when the interference has a frequency fluctuation, the ideal adaptive-filter coefficients for cancelling the interference also fluctuate. When the adaptation property of the algorithm is slow compared with the frequency fluctuation, the interference-cancelling performance is degraded. However, if the adaptation is too quick, the performance is degraded owing to the target signal. To overcome this problem, we propose an adaptive filter that suppresses the fluctuation of the ideal coefficients by utilizing a $\frac{\pi}{2}$ phase-delay device. This method assumes a frequency response that characterizes the transmission path from the interference source to the main input signal to be sufficiently smooth. In the numerical examples, the proposed method exhibits good performance in the presence of a frequency fluctuation when the forgetting factor is large. Moreover, we show that the proposed method reduces the calculation cost.

Purchase behavior prediction is one of the most important issues to promote both e-commerce companies' sales and the consumers' satisfaction. The prediction usually uses features based on the statistics of items. This kind of features can lead to the loss of detailed information of items. While all items are included, a large number of features has the negative impact on the efficiency of learning the predictive model. In this study, we propose to use the most popular items for improving the prediction. Experiments on the real-world dataset have demonstrated the effectiveness and the efficiency of our proposed method. We also analyze the reason for the performance of the most popular items. In addition, our work also reveals if interactions among most popular items are taken into account, the further significant improvement can be achieved. One possible explanation is that online retailers usually use a variety of sales promotion methods and the interactions can help to predict the purchase behavior.

With the popularization of Internet of things (IoT), the interaction between human and IoT has become a daily life. In this interaction, the objects of IoT usually require access to personal data, which are often sensitive. We propose a lightweight privacy-preserving model based on the trust evaluation that it can effectively protect privacy based on simple threshold detection. The key issue we address in this work is how to construct trust model so that non trusted objects were prevented from accessing private data. This work can be considered as a lightweight approach to access control for privacy-preservation. The main algorithm in the proposed model is a kind of dynamic self-adjusting trust evaluation mechanism that uses a combination of interaction information occurs between the human and the Internet of things, between the human and the human. According to the given threshold, the trust model can determine the data level of object access in the IoT. We have implemented a prototype of the proposed scheme, thereby demonstrating the feasibility of the proposed scheme on resource-constrained devices.

Network coding (NC) is considered a new paradigm for distributed networks. However, NC has an all-or-nothing property. In this paper, we propose a sparse recovery approach using sparse sensing matrix to solve the NC all-or-nothing problem over a finite field. The effectiveness of the proposed approach is evaluated based on a sensor network.

The study focuses on the adaptation problem for HTTP low-delay live streaming over mobile networks. In this context, the client's small buffer could be easily underflown due to throughput variations. To maintain seamless streaming, we present a probabilistic approach to adaptively decide the bitrate for each video segment by taking into account the instant buffer level. The experimental results show that the proposed method can significantly reduce buffer underflows while providing high video bitrates.

In this letter, we propose a novel method for face hallucination by learning a new distance metric in the low-resolution (LR) patch space (source space). Local patch-based face hallucination methods usually assume that the two manifolds formed by LR and high-resolution (HR) image patches have similar local geometry. However, this assumption does not hold well in practice. Motivated by metric learning in machine learning, we propose to learn a new distance metric in the source space, under the supervision of the true local geometry in the target space (HR patch space). The learned new metric gives more freedom to the presentation of local geometry in the source space, and thus the local geometries of source and target space turn to be more consistent. Experiments conducted on two datasets demonstrate that the proposed method is superior to the state-of-the-art face hallucination and image super-resolution (SR) methods.

Object detection in millimeter-wave Interferometric Synthetic Aperture Radiometer (InSAR) imaging is always a crucial task. Facing unpredictable and numerous objects, traditional object detection models running after the InSAR system accomplishing imaging suffer from disadvantages such as complex clutter backgrounds, weak intensity of objects, Gibbs ringing, which makes a general purpose saliency detection system for InSAR necessary. This letter proposes a spectrum-based saliency detection algorithm to extract the salient regions from unknown backgrounds cooperating with sparse sensing InSAR imaging procedure. Directly using the interferometric value and sparse information of scenes in the basis of the Discrete Cosine Transform (DCT) domain adopted by InSAR imaging procedure, the proposed algorithm isolates the support of saliency region and then inversely transforms it back to calculate the saliency map. Comparing with other detecting algorithms which run after accomplishing imaging, the proposed algorithm will not be affected by information-loss accused by imaging procedure. Experimental results prove that it is effective and adaptable for millimeter-wave InSAR imaging.

Building detection from high resolution remote sensing images is challenging due to the high intraclass variability and the difficulty in describing buildings. To address the above difficulties, a novel approach is proposed based on the combination of shape-specific feature extraction and discriminative feature classification. Shape-specific feature can capture complex shapes and structures of buildings. Discriminative feature classification is effective in reflecting similarities among buildings and differences between buildings and backgrounds. Experiments demonstrate the effectiveness of the proposed approach.

Shadow is an important effect that makes virtual 3D scenes more realistic. In this paper, we propose a fast and correct soft shadow generation method for area lights of various shapes and colors. To conduct efficient as well as accurate visibility tests, we exploit the complexity of shadow and area light color.

Due to heavy rendering load and unstable frame rate when rendering large terrain, this paper proposes a geometry clipmaps based algorithm. Triangle meshes are generated by few tessellation control points in GPU tessellation shader. ‘Cracks’ caused by different resolution between adjacent levels are eliminated by modifying outer tessellation level factor of shared edges between levels. Experimental results show the algorithm is able to improve rendering efficiency and frame rate stability in terrain navigation.

In this paper, we propose a method to find similar sentences based on language resources for building a parallel corpus between English and Korean from Wikipedia. We use a Wiki-dictionary consisted of document titles from the Wikipedia and bilingual example sentence pairs from Web dictionary instead of traditional machine readable dictionary. In this way, we perform similarity calculation between sentences using sequential matching of the language resources, and evaluate the extracted parallel sentences. In the experiments, the proposed parallel sentences extraction method finally shows 65.4% of F1-score.

This paper presents human-centered video feature selection via mRMR-SCMMCCA (minimum Redundancy and Maximum Relevance-Specific Correlation Maximization Multiset Canonical Correlation Analysis) algorithm for preference extraction. The proposed method derives SCMMCCA, which simultaneously maximizes two kinds of correlations, correlation between video features and users' viewing behavior features and correlation between video features and their corresponding rating scores. By monitoring the derived correlations, the selection of the optimal video features that represent users' individual preference becomes feasible.