Problem In Many Applications Research Articles

Joint Forecasting and Interpolation of Time-Varying Graph Signals Using Deep Learning

We tackle the problem of forecasting network-signal snapshots using past signal measurements acquired by a subset of network nodes. This task can be seen as a combination of multivariate time-series forecasting (temporal prediction) and graph-signal interpolation (spatial prediction). This is a fundamental problem for many applications wherein deploying a high granularity network is impractical. Our solution combines recurrent neural networks with frequency-analysis tools from graph signal processing, and assumes that data is sufficiently smooth with respect to the underlying graph. The proposed learning model outperforms state-of-the-art deep learning techniques, especially when predictions are made using a small subset of network nodes, considering two distinct real world datasets: temperatures in the US and speed flow in Seattle. The results also indicate that our method can handle noisy signals and missing data, making it suitable to many practical applications.

On the Use of Biased-Randomized Algorithms for Solving Non-Smooth Optimization Problems

Soft constraints are quite common in real-life applications. For example, in freight transportation, the fleet size can be enlarged by outsourcing part of the distribution service and some deliveries to customers can be postponed as well; in inventory management, it is possible to consider stock-outs generated by unexpected demands; and in manufacturing processes and project management, it is frequent that some deadlines cannot be met due to delays in critical steps of the supply chain. However, capacity-, size-, and time-related limitations are included in many optimization problems as hard constraints, while it would be usually more realistic to consider them as soft ones, i.e., they can be violated to some extent by incurring a penalty cost. Most of the times, this penalty cost will be nonlinear and even noncontinuous, which might transform the objective function into a non-smooth one. Despite its many practical applications, non-smooth optimization problems are quite challenging, especially when the underlying optimization problem is NP-hard in nature. In this paper, we propose the use of biased-randomized algorithms as an effective methodology to cope with NP-hard and non-smooth optimization problems in many practical applications. Biased-randomized algorithms extend constructive heuristics by introducing a nonuniform randomization pattern into them. Hence, they can be used to explore promising areas of the solution space without the limitations of gradient-based approaches, which assume the existence of smooth objective functions. Moreover, biased-randomized algorithms can be easily parallelized, thus employing short computing times while exploring a large number of promising regions. This paper discusses these concepts in detail, reviews existing work in different application areas, and highlights current trends and open research lines.

APPLICATION OF B-SPLINE METHOD IN SURFACE FITTING PROBLEM

Abstract. Fitting a smooth surface on irregular data is a problem in many applications of data analysis. Spline polynomials in different orders have been used for interpolation and approximation in one or two-dimensional space in many researches. These polynomials can be made by different degrees and they have continuous derivative at the boundaries. The advantage of using B-spline basis functions for obtaining spline polynomials is that they impose the continuity constraints in an implicit form and, more importantly, their calculation is much simpler. In this study, we explain the theory of the least squares B-spline method in surface approximation. Furthermore, we present numerical examples to show the efficiency of the method in linear, quadratic and cubic forms and it’s capability in modeling changes in numerical values. This capability can be used in different applications to represent any natural phenomenon which can’t be experienced by humans directly. Lastly, the method’s accuracy and reliability in different orders will be discussed.

Research on Author Name Disambiguation Based on Fusion Features and Semantic Fingerprints

Author name disambiguation has been a challenging problem in many applications. In order to promote researches to solve name disambiguation, Aminer launched the Open Academic Data Challenge 2018 jointly with Chinese Association for Artificial Intelligence and China Knowledge Centre for Engineering and Technology. Aminer is a scholar-cantered academic search and mining platform covering more than 200 million papers and more than 100 million scholars in various academic fields. Our team proposed a name disambiguation method based on fusion features and semantic fingerprint technique to participate in the competition. The method identified authors with same names through organization feature and co-author feature at first, and then it solves ambiguity names by way of semantic fingerprints which are 128-bit binary vector generated from textual features of papers by Simhash algorithm. Our method scored 0.609 on the verification set and 0.879 on the test set ranking first in the final submission.

Nanocomposite coatings on steel for enhancing the corrosion resistance: A review

Steel is known for its low cost of fabrication, high mechanical strength and hence is extensively used for drilling equipment, pipelines, ship building and offshore structures. Corrosion of steel is a costly problem in many applications especially in oilfield and marine environments which are known for the high temperature, high pressure and corrosive conditions. In this paper, nanocomposite coating is being explored as the preferred strategy to improve corrosion resistance for steel. Here, we will give details on the various coating materials, deposition techniques and the challenges involved in realising the most suitable coating on steel based on results of recent research. In addition, we also detail the filler specifications for getting high performance nanocomposites.

Optimization of multitask parallel mobile edge computing strategy based on deep learning architecture

As a mainstream computing and storage strategy for mobile communications, Internet of Things and large data applications, mobile edge computing strategy mainly benefits from the deployment and allocation of small base stations. Mobile edge computing mainly helps users to complete complex, intensive and sensitive computing tasks. However, the algorithm has many problems in practical application, such as complex user needs, complex user mobility, numerous services and applications. Therefore, under the above background, it is of great significance to solve the computational pressure of current mobile edge algorithm and optimize its algorithm architecture. This paper creatively proposes a deep learning architecture based on tightly connected network, and transplants it into mobile edge algorithm to realize the payload sharing process of edge computing, so as to establish an efficient network model. At the same time, we creatively propose a multi-task parallel scheduling algorithm, which realizes the mobile edge algorithm in the face of complex computing and algorithm efficiency. Finally, the above algorithms are simulated and tested. The experimental results show that under the same task, the time consumed by the proposed algorithm is 3.5–4, while the time consumed by the traditional algorithm is 4.5–8, and the corresponding time is standardized time, so the practice shows that the algorithm has obvious overall efficiency advantages.

Investigation of incremental hybrid genetic algorithm with subgraph isomorphism problem

Graph pattern matching is a key problem in many applications which data is represented in the form of a graph, and this problem is generally defined as a subgraph isomorphism. In this paper, we analyze an incremental hybrid genetic algorithm for the subgraph isomorphism problem considering various design issues to improve the performance of the algorithm. An incremental hybrid genetic algorithm was previously suggested to solve the subgraph isomorphism problem and have shown good performance. It decomposes the problem into a sequence of consecutive subproblems which has an optimal substructure. Each subproblem is solved by the hybrid genetic algorithm and the solutions obtained are extended to be applied to the next subproblem as initial solutions. We examine a wide range of schemes that determine the overall performance of the incremental process and make a number of experiments to verify the effectiveness of each scheme with the synthetic dataset of random graphs. We show that the performance of incremental approach can be significantly improved compared to the previous representative studies by applying appropriate schemes found by the experiments. In addition, we also investigate the effect of different genetic parameters and identify the scalability of our method by conducting experiments using real world dataset with large-sized graphs.

Foreground Detection for Infrared Videos With Multiscale 3-D Fully Convolutional Network

Foreground detection for infrared (IR) videos is an important and fundamental problem in many applications, e.g., IR surveillance, IR object tracking, and so on. Conventional foreground detection algorithms developed for visible videos do not focus on the problems for IR videos, e.g., low contrast, coarse texture, lack of color information, and so on. Recent foreground detection methods based on deep neural network (DNN) demonstrated significant improvement, but most of them still use only spatial features, which is less obvious in IR images. In this letter, we add deeply learned multiscale temporal features to improve the performance of background subtraction for IR videos. We propose a novel multiscale 3-D fully convolutional network (MFC3-D) to establish a mapping from image sequences to pixelwise classification results and to learn deep and hierarchical multiscale spatial–temporal features of the input images sequence. The experimental results show that the MFC3-D can learn spatial–temporal features effectively and achieved state-of-the-art results on the test data set, comparing to other DNN-based methods and traditional background subtraction methods.

Fast and scalable algorithm for product large data on multicore system

SummaryThe problem of designing efficient parallel algorithms to calculate the product of n numbers when the multipliers are large is a fundamental problem in many applications of computer science such as cryptography. In this work, we present a new parallel algorithm on exclusive read shared memory model. The performance of the introduced algorithm is measured based on three factors, namely, (1) the number of cores, (2) the size of the array, and (3) the size of the multiplier. The experimental study on a multi core system reveals that the introduced algorithm is faster than the best‐known optimal parallel algorithm. The improvement of the proposed algorithm in processing time compared to the best known parallel algorithm is 80% when the size of the array was 220 and the sizes of the multiplier were 1024, 2048, and 4096 bits. Moreover, our algorithm is a highly scalable parallel algorithm compared with the best‐known optimal parallel algorithm.

People Identification Based on Person Image and Additional Physical Parameters Comparison

This paper proposes and presents one approach for people identification based on image and additional physical parameters, height and step length, of a person comparison. People identification is very important in many areas of human life. There are large number of identification methods (biometric methods) that include a different scope of methods, for example fingerprint identification, hand geometry identification, facial recognition, methods based on human eye identification (retina and iris), gait recognition etc. Most of that methods require some kind of interaction with a person, what could be a problem in many practical applications. The method that does not require any interaction with a person is gait recognition. One approach for a people identification based on gait recognition, that uses silhouettes of a person and parameters of person height and step length, is proposed and presented in this paper.

Edge Computing: Applications, State-of-the-Art and Challenges

The Internet of Things (IoT) is now infiltrating into our daily lives, providing important measurement and collection tools to inform us of every decision. Millions of sensors and devices continue to generate data and exchange important information through complex networks that support machine-to-machine communication and monitor and control critical smart world infrastructure. As a strategy to alleviate resource congestion escalation, edge computing has become a new paradigm for addressing the needs of the Internet of Things and localization computing. Compared to well-known cloud computing, edge computing migrates data calculations or storage to the edge of the network near the end-user. Thus, multiple compute nodes distributed across the network can offload computational pressure from a centralized data center and can significantly reduce latency in message exchanges. Besides, the distributed architecture balances network traffic and avoids spikes in traffic in the IoT network, reduces latency between edge/cloud servers and end-users, and reduces response time for real-time IoT applications compared to traditional cloud services. In this article, we conducted a comprehensive survey to analyze how edge computing can improve the performance of IoT networks. We classify edge calculations into different groups based on the architecture and study their performance by comparing network latency, bandwidth usage, power consumption, and overhead. Through the systematic introduction of the concept of edge computing, typical application scenarios, research status, and key technologies, it is considered that the development of edge computing is still in the initial stage. There are still many problems in practical applications that need to be solved, including optimizing edge computing performance, security, interoperability, and intelligent edge operations management services.

Предсказание эволюции динамических систем остаточными нейронными сетями

Forecasting time series and time-dependent data is a common problem in many applications. One typical example is solving ordinary differential equation (ODE) systems $\dot{x}=F(x)$. Oftentimes the right hand side function $F(x)$ is not known explicitly and the ODE system is described by solution samples taken at some time points. Hence, ODE solvers cannot be used. In this paper, a data-driven approach to learning the evolution of dynamical systems is considered. We show how by training neural networks with ResNet-like architecture on the solution samples, models can be developed to predict the ODE system solution further in time. By evaluating the proposed approaches on three test ODE systems, we demonstrate that the neural network models are able to reproduce the main dynamics of the systems qualitatively well. Moreover, the predicted solution remains stable for much longer times than for other currently known models.

Generalizing the Pigeonhole Principle for Similarity Search in Hamming Space

A distance search in Hamming space finds binary vectors whose Hamming distances are no more than a threshold from a query vector. It is a fundamental problem in many applications, such as image retrieval, near-duplicate Web page detection, and scientific databases. State-of-the-art approaches to Hamming distance search are mainly based on the pigeonhole principle to generate a set of candidates and then verify them. We observe that the constraint by the pigeonhole principle is not always tight and may bring about unnecessary candidates. We also observe that the distribution in real data is often skewed, but most existing solutions adopt a simple equi-width partitioning and allocate the same threshold to all the parts, hence failing to exploit the data skewness to optimize query processing. In this paper, we propose a new form of the pigeonhole principle which allows variable partitioning and threshold allocation. Based on the new principle, we develop a tight constraint of candidates and devise cost-aware methods for partitioning and threshold allocation to optimize query processing. In addition, we extend our methods to answer Hamming distance join queries. We also discuss the application of the pigeonhole principle in set similarity search, a problem that can be converted to Hamming distance search equivalently. Our evaluation on datasets with various data distributions shows the robustness of our solution and its superior query processing performance to the state-of-the-art methods.

A Novel MPPT Method Based on Cuckoo Search Algorithm and Golden Section Search Algorithm for Partially Shaded PV System

Partial shading is a common and difficult problem to be solved in a photovoltaic (PV) system. Numerous efforts have been introduced to mitigate this problem. Some commonly used approaches are deploying some metaheuristic (MH) algorithm to track the multiple-peak P-V curve of partially shaded PV system. Cuckoo search (CS) is a new optimization algorithm based on the MH approach. It has been used to solve an optimization problem in many applications, including the maximum power point tracking (MPPT) problem. The CS algorithm performs well in tracking the global maximum power point (GMPP). However, just like any other MH algorithm, there is still a dilemmatic trading between their accuracy and the tracking time needed to find GMPP. This paper proposes a new MPPT algorithm by combining the CS algorithm with golden section search (GSS) to take beneficial features from both the algorithms. To validate the proposed algorithm, it is evaluated with various cases of partial shading. The simulation and experimental results show a noticeable performance improvement compared with the original CS algorithm and other MH algorithms.

Construction of wavelets and framelets on a bounded interval

Many problems in applications are defined on a bounded interval. Therefore, wavelets and framelets on a bounded interval are of importance in both theory and application. There is a great deal of effort in the literature on constructing various wavelets on a bounded interval and exploring their applications in areas such as numerical mathematics and signal processing. However, many papers on this topic mainly deal with individual examples which often have many boundary wavelets with complicated structures. In this paper, we shall propose a method for constructing wavelets and framelets in [Formula: see text] from symmetric wavelets and framelets on the real line. The constructed wavelets and framelets in [Formula: see text] often have a few simple boundary wavelets/framelets with the additional flexibility to satisfy various desired boundary conditions. To illustrate our construction method, from several spline refinable vector functions, we present several examples of (bi)orthogonal wavelets and spline tight framelets in [Formula: see text] with very simple boundary wavelets/framelets.

An Operational Concept for Correcting Navigation Drift During Sonar Surveys of the Seafloor

The accumulation of navigation errors (drift) is a problem in many applications of autonomous underwater vehicles (AUVs), particularly during long-duration underwater surveys. Traditional methods for correcting drift require either surfacing of the vehicle for a global navigation satellite systemupdate or use of an independent acoustic positioning system. These methods may not be desirable or possible due to mission constraints. We propose a solution to this problem completely underwater and without the aid of external navigation systems. The approach is based on an operational concept that uses a modified paired-track survey pattern combined with through-the-sensor navigation corrections from a seafloor imaging sonar. We describe the operational concept, derive a model for its performance limits, validate this model, and demonstrate the concept with real experiments at sea. Using this approach, we provide an opportunity to use either coherent or incoherent through-the-sensor positioning corrections for a mission length increase of only the product of the intratrack spacing and the number of track pairs. We show results from a proof-of-principle experiment using data collected by the ${\text{300-kHz}}$ synthetic aperture sonar of the NATO Centre for Maritime Research and Experimentation's Minehunting Unmanned underwater vehicle for Shallow water Covert Littoral Expeditions.

Modal energy transfer by thermally induced refractive index gratings in Yb-doped fibers

Thermally induced refractive index gratings in Yb-doped fibers lead to transverse mode instability (TMI) above an average power threshold, which represents a severe problem for many applications. To obtain a deeper understanding of TMI, the evolution of the strength of the thermally induced refractive index grating with the average output power in a fiber amplifier is experimentally investigated for the first time. This investigation is performed by introducing a phase shift between the refractive index grating and modal interference pattern, which is obtained by applying a pump power variation to the fiber amplifier. It is demonstrated that the refractive index grating is sufficiently strong to enable modal energy coupling at powers that are significantly below the TMI threshold if the induced phase shift is sufficiently large. The experiments indicate that at higher powers, the refractive index grating becomes more sensitive to such phase shifts, which will ultimately trigger TMI. Furthermore, the experimental results demonstrate beam cleaning above the TMI threshold via the introduction of a positive phase shift. This finding paves the way for the development of a new class of mitigation strategies for TMI that are based on controlling the phase shift between the thermally induced refractive index grating and modal interference pattern.

A new similarity/distance measure between intuitionistic fuzzy sets based on the transformed isosceles triangles and its applications to pattern recognition

Intuitionistic fuzzy set (IFS) is a representative model of fuzzy theory, and it is widely used to deal with fuzzy and uncertain problems in many practical applications. In IFS theory, the calculation of the distance and similarity between IFSs are significant techniques that are effective measurement methods for distinguishing the similarity degree between IFSs, and the two measures can be used in pattern recognition, decision-making and so on. In this paper, a novel similarity/distance measure between IFSs is proposed according to the intersections of the transformed isosceles triangles from IFSs, and the isosceles triangles are placed in a square area; furthermore, the properties of the proposed measure are also analyzed to prove whether the measure satisfies the definition of the similarity/distance measure for IFSs or not. The numerical experiments are implemented to test the validity of the proposed measure; in addition, several pattern recognition and clustering problems are also employed to further demonstrate its effectiveness. The experimental results show that the proposed measure is an accurate and superior measure that can avoid the shortcomings of most existing measures. Finally, the directions for future research of this work are also represented in the conclusion.

Chaotic dragonfly algorithm: an improved metaheuristic algorithm for feature selection

Selecting the most discriminative features is a challenging problem in many applications. Bio-inspired optimization algorithms have been widely applied to solve many optimization problems including the feature selection problem. In this paper, the most discriminating features were selected by a new Chaotic Dragonfly Algorithm (CDA) where chaotic maps embedded with searching iterations of the Dragonfly Algorithm (DA). Ten chaotic maps were employed to adjust the main parameters of dragonflies’ movements through the optimization process to accelerate the convergence rate and improve the efficiency of DA. The proposed algorithm is employed for selecting features from the dataset that were extracted from the Drug bank database, which contained 6712 drugs. In this paper, 553 drugs that were bio-transformed into liver are used. This data have four toxic effects, namely, irritant, mutagenic, reproductive, and tumorigenic effect, where each drug is represented by 31 chemical descriptors. The proposed model is mainly comprised of three phases; data pre-processing, features selection, and the classification phase. In the data pre-processing phase, Synthetic Minority Over-sampling Technique (SMOTE) was used to solve the problem of the imbalanced dataset. At the features selection phase, the most discriminating features were selected using CDA. Finally, the selected features from CDA were used to feed Support Vector Machine (SVM) classifier at the classification phase. Experimental results proved the capability of CDA to find the optimal feature subset, which maximizing the classification performance and minimizing the number of selected features compared with DA and the other meta-heuristic optimization algorithms. Moreover, the experiments showed that Gauss chaotic map was the appropriate map to significantly boost the performance of DA. Additionally, the high obtained value of accuracy (81.82–96.08%), recall (80.84–96.11%), precision (81.45–96.08%) and F-Score (81.14–96.1%) for all toxic effects proved the robustness of the proposed model.

Fast algorithms for sparse inverse covariance estimation

ABSTRACTSparse precision matrix (i.e. inverse covariance matrix in statistic term) estimation is an important problem in many applications of multivariate analysis. The problem becomes very challenging when the dimension of data is much larger than the number of samples. In this paper, we propose a convex relaxation model for the sparse covariance selection problem, which is solved by the well-known alternating direction method of multipliers (ADMM). Furthermore, a new model with positive semi-definite constraint is proposed. Numerical results show that the ADMM-based methods perform favourably compared with the column-wise manner on both synthetic and real data.