Introduction. Graphene coated magnetic nanoparticles (GCMNP) are object of a lot a research addressed to improve chemical and thermal stability and biocompatibility of magnetic nanoparticles (MNPs), in view of the exploitation of their properties in different applications including catalytic, environmental, biological, biomedical and electronic. Carbon covering has many advantages over other coatings, such as much higher chemical and thermal stability, easy functionalization. Typically, a two step process (MNP synthesis and then coating) is performed, however producing GCMNPs in a single step is a fascinating challenge [1]. Among different strategies suggested to prepare MNPs, Chemical Vapor Deposition (CVD) is the easiest one to be scaled up towards an economically viable production [2]. In fact, little attention has been devoted to the effect of process parameters in the preparation of stable MNPs, to obtain a quality controlled product and this is even more rare for GCMNPs. For a given catalyst and carbon source the CVD products strongly depend of the operating conditions and the selective and controlled coating process is still to be understood and optimized. Moreover, little attention has been devoted to investigate the influence of the support on the GCMNPs characteristics. Finally, insight into the formation mechanism is a critical issue to improve the control of the synthesis process. Herein, we report the preparation of stable core-shell graphene-coated magnetic nanoparticles (GCMNPs) via Catalytic Chemical Vapor Deposition (CCVD) of methane at atmospheric pressure. The magnetic properties of the nanoparticles have been also investigated. Materials and methods. The Co, Fe catalyst (50 wt.% of each metal) was prepared by wet impregnation of gibbsite (γ-Al(OH)3) powder [3]. The experimental plant for the synthesis was equipped with on-line analyzers (Uras 26, ABB) that permit the monitoring of the inlet and outlet reactor concentrations of the reactants. To characterize the reaction products various techniques were employed as follows: transmission electron microscopy (TEM) (FEI Tecnai electron microscope operating at 200 kV), scanning electron microscopy (SEM) (LEO 1525 microscope), Raman spectroscopy ((Renishaw inVia; 514 nm excitation wavelength), thermogravimetric analysis (TG-DTG) (SDTQ 600 Analyzer (TA Instruments)) coupled with a quadrupole mass detector, X-ray diffraction analysis (Bruker D8 X-ray diffractometer) and N2 adsorption–desorption at 77 K. Results The change of process parameters: total flow rate, hydrocarbon methane partial pressure and catalyst weight in the synthesis process, has shown that the selective covering of nanoparticles and the control of the coating thickness can be obtained by feeding the hydrocarbon in a suitable carrier, preventing the unwanted homogeneous decomposition and increasing the conversion of methane.
As more and more data is outsourced to cloud which is assumed to be a semi-trusted server, it is necessary to encrypt the sensitive data stored in the cloud. However, it brings a series of problems, such as: How to search over the encrypted data efficiently and securely? How should a data owner grant search capabilities to the data users? To solve these problems, we propose two attribute-based keyword search and data access control schemes based on public-key searchable encryption and attribute based encryption. Our solutions allow a data owner to control the access policy and grant the search policy to any data user who wants to retrieve the encrypted data efficiently.
For multiview face retrieval of certain person in surveillance video, a key challenge is the lack of training samples. Generally, the law enforcement agencies usually have only one frontal view face image of the target person, however, the faces of the target person in the surveillance video could be in different orientation, and it is impossible for a classifier trained on only frontal view face to retrieve the faces under other orientation. This paper proposes an active training sample collection method for multiview face retrieval in surveillance video. First, the front view face image is used to train a classifier to retrieve the target person's front view face in videos. As the video is continuous, we can track the face and obtain side view faces of the target person. Then these selected side view faces are combined with the frontal view face to form a new training data set. The classifier is updated based on the new training data set, and can retrieve multiview faces of the target people. The experimental results prove the effectiveness of the proposed method.
Cloud computing technology is a relatively new concept of providing dramatically scalable and virtualized resources, software and hardware on demand to consumers. Cloud Computing offers a whole new paradigm to provide users with high end and scalable infrastructure at an affordable cost. It is based on many new technologies like virtualization and distributed computing. However, security concerns are terrible for these systems whose infrastructure and computational resources are owned by an outside party that sells those services to the general public. In fact, data breaches to cloud services are also increasing every year due to hackers who are always trying to exploit the security vulnerabilities of the cloud architecture. We present in this paper a detailed analysis of the cloud computing security issues and the general security requirements into which the security concerns falls. In addition, countermeasures to cloud security breaches are proposed.
The traditional financial time series forecasting methods use accurate input data for prediction, and then make single-step or multi-step prediction based on the established regression model. So its prediction result is one or more specific values. But because of the complexity of financial markets, the traditional forecasting methods are less reliable. In this paper, we transform the financial time series into fuzzy grain particle sequences, and use support vector machine regression to regress the upper and lower bounds of the fuzzy particles, and then apply regression model single-step prediction on the upper and lower bounds, which will limit the predict results within a range. This is a new idea. The Shanghai Composite Index Week closed index for the experimental data, experimental results show the effectiveness of this approach.
Query is one of the most important factors that can directly influence the results of information retrieval (IR). However, the query is defined by the user and thus inevitably has the following two problems: (1) the user often cannot exactly represent their search intention via query terms, (2) the user cannot effectively select the weight of each query term based on its importance toward the query's meaning. The above two problems cause the two types of uncertainty of a query. In this paper, we define them as the uncertainty of the query structure and the uncertainty of the query parameter, respectively. To eliminate the above two types of uncertainty and solve the above two problems, this paper proposes a new algorithm which includes two parts: (1) a self-organizing query structure loop which expands the initial query by adding only one term within each loop based on feedback technology until it meets the terminating condition of expansion defined by the author, and (2) an optimization algorithm based on a genetic algorithm (GA) that optimizes the weights of the expanded query vector within each loop. This algorithm provides a method of finding the optimal number of query expansion terms and improving the precision and recall of the search results. The experiment results show the effectiveness of the proposed algorithm.
The particle swarm optimizer algorithm is a bio-inspired optimization principle and a typical swarm intelligence algorithm whose theory base is random theory. Many researchers attempted to make the PSO process clearly from the perspective of the random theory but got some complicated and nonobjective results. In this paper from the perspective of base theory, the fitness function that evaluates the performance of particles in the swarm is shown by a fuzzy random variable while the convergence and its speed of PSO process can be shown by the two parameters (the belief level value and the Borel set) of the chance measure of fuzzy random variable. Then we can obtain some straightforward and concrete results.
In this paper, we propose a novel random sampling algorithm for the shortest vector problem (SVP) based on the y-sparse representations of the short lattice vectors. The experimental results show that the random sampling algorithm outperforms the other two SVP algorithms under the benchmarks of SVP challenge[1]. Therefore, the random sampling algorithm is an efficient SVP solver for the shortest vector problem.
This paper presents a classify model of learning behavior on an intelligent tutoring website which provides a describing method of learning behavior. MFR Model creates a conceptual model and a analysis model according to students' learning behavior on the website. By using MFR model, influences of participation and learning effects can be judged by learning behavior which can not only strengthen the ability of interaction but also improve teaching quality and efficiency by providing personalized teaching method on website.
The ground effect of aircraft has been well studied in recent decades. However, available studies (inclusive of test techniques) are all carried out based on quasi-steady methods, without considering the effects of the time course on the flow field. That is, the status of every time point is independent from each other, and the flow field of the current point has no effect on its next one. In this paper, the problem of unsteady ground effect of flying wing UAV is solved by using the computational fluid dynamics (CFD) method, where the dynamic laying method is adopted to generate moving grid caused by the relative motion of aircraft and aircraft carrier deck, and Euler equation is calculated to simulate the unsteady flow field. Finally, computational results show that the force and moment under unsteady conditions is more close to the real physical phenomenon while comparing with quasi-steady methods.
Brain storm optimization is a new swarm intelligence, which mimics the human brainstorming process. In this paper, a modified brain storm optimization is proposed based on uncertainty information. It adopts affinity propagation clustering instead of k-means clustering. Meanwhile, a creating operator combining the information of multiple clusters is introduced by borrowing the idea of cloud drops algorithm. The proposed brain storm optimization is characterized by mining and utilizing the uncertain information of candidate solutions with no need for the number of clusters. Finally, the modified brain storm optimization is applied to numerical optimization. The simulation results show that the proposed algorithm has better optimization results and higher rate of success than the original version.
Packet filter system based on high speed match engine of REGular EXPressions (REGEXP) plays an important role in domain of Intrusion Detection System (IDS), Deep Packet Inspection (DPI) system, network security and traffic monitoring, etc. However, the existing filter schemas suffer from several deficiencies in matching speed and memory footprint, such as traditional DFA matching, single-level signature hash and DFA grouping. To overcome these shortcomings, in this paper, a new packet filter schema based on multilevel signature and DFA grouping is proposed. In particular, an algorithm called "DFA pseudo-split" is presented in our proposal to overcome the shortage of signatures. The experimental results show that our proposal significantly outperforms the traditional filter schemas.
An improved adaptive median filter is proposed to remove the salt and pepper noise. The algorithm determines a pixel point as signal pixel or possible noise pixel according to the characteristic of the salt and pepper noise, the noise pixel is removed by adaptive median filter, the signal pixel is kept. The results show that the algorithm effectively removes the salt and pepper noise while protecting the edge details of the image.
This paper gives an empirical study on determinants of capital structure of Chinese-listed companies using firm-level panel data. The study employs a new database containing the accounting data from 89 non-financial companies listed on the Shanghai and Shenzhen Stock Exchange during the period time of 2003 to 2005. It employs six index variables as independent variables, and finds main affecting factors by multiple regression analysis. It is concluded that the long-term debt ratio positively correlated with tangibility and growth opportunities, however, the size, profitability and non-debt tax shield have a negative impact on the long-term debt ratio. Overall, most results of this study are consistent with the propositions that based on the research about other countries, which implies that the Chinese listed firms have followed the basic rules of global economy despite the state controlling ownership. Another empirical finding shows that Chinese firms tend to have lower long-term debt compare with other countries.
We propose a simple but effective tracking algorithm for non-rigid objects with geometric appearance changes. The discriminative features of the object are adaptively selected according to their descriptive ability. To adapt to the geometric changes, we use a deformable rectangle to represent the object, and use Markov Chain Monte Carlo-based Particle Filter (MCMC-PF) to estimate the state of the object in a restricted four dimensional space. Experimental results show that the proposed tracking algorithm has ideal performance.
