Search

Support vector machine (SVM) is one of the most powerful technologies of machine learning, which has been widely concerned because of its remarkable performance. However, when dealing with the classification problem of large-scale datasets, the high complexity of SVM model leads to low efficiency and become impractical. Due to the sparsity of SVM in the sample space, this paper presents a new parallel data geometry analysis (PDGA) algorithm to reduce the training set of SVM, which helps to improve the efficiency of SVM training. The PDGA introduce Mahalanobis distance to measure the distance from each sample to its centroid. And based on this, proposes a method that can identify non support vectors and outliers at the same time to help remove redundant data. When the training set is further reduced, cosine angle distance analysis method is proposed to determine whether the samples are redundant data, ensure that the valuable data are not removed. Different from the previous data geometry analysis methods, the PDGA algorithm is implemented in parallel, which greatly saving the computational cost. Experimental results on artificial dataset and 6 real datasets show that the algorithm can adapt to different sample distributions. Which significantly reduce the training time and memory requirements without sacrificing the classification accuracy, and its performance is obviously better than the other five competitive algorithms.

Polyvinyl alcohol- (PVA-) based aerogels have attracted widespread attention owing to their low cost, eco-friendliness, and low density. However, the applications of PVA-based aerogels are limited by their flammability. In this study, a flame retardant, ammonium polyphosphate (APP), and a biopolymer, chitosan (CS), were added to polyvinyl alcohol (PVA), and the polymer was further crosslinked using boric acid (H3BO3). In the PVA aerogels, the negatively charged APP and positively charged CS formed a polyelectrolyte complex (PEC) through ionic interaction. Cone calorimetry and vertical burning tests (UL-94) indicated that the PVA composite aerogels have excellent flame retardancy; they could decrease the heat release rate, total heat release rate, and carbon dioxide (CO2) generation. Both PVA/H3BO3 and APP-CS in the composite aerogel could be burned to carbon, and the foamed char layer could act together to impart the PVA composite aerogels with good flame retardancy. Further, the decrease in the temperature at the backside of the aerogels with increasing APP-CS content, as determined by the flame-spraying experiment, indicated that the PVA-based aerogels with APP-CS can also serve as thermal insulation materials. This work provides an effective and promising method for the preparation of PVA-based aerogels with good flame retardancy and thermal insulation property for construction materials.

In this work, a new synergistic charring agent of melamine cyanurate-fumed silica (MCA-SiO2) has been synthesized through the facile self-assembly process, where melamine (ME) and cyanuric acid (CA) reacted and deposited on the fumed silica in the aqueous suspension. Then GF-PP/IFR-(MCA-SiO2) composites were prepared by melting blend of GF-PP and IFR containing MCA-SiO2. The general properties of the resulting composites were characterized by Cone calorimeter test (CCT), scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FT-IR) etc. Results show that the addition of SiO2 could greatly reduced the heat release rate (HRR) of the GF-PP/IFR-MCA composites and improved the flame retardant properties during burning. When the content of SiO2 in MCA-SiO2 was 20 wt%, the flame retardant property of the composites reached the UL-94 V-0 rating, corresponding to a the limiting oxygen index (LOI) of 32.4 %. This enhanced flame retardant property could be attributed to the fact that the fumed silica with low thermal conductivity could prevent the heat exchange, and assist in the formation of compact and dense char during the burning.

Text classification is the critical content of machine learning, and it is widely applied in information filtering, sentimental analysis, and text review. It is very important to improve the accuracy of classification results, and this is also the main research purpose of researchers in this field in recent years. Feature selection plays an important role in text classification, which has the functions of eliminating irrelevant features, reducing dimensionality, and improving classification accuracy. So, this paper studies the CHI feature selection algorithm, and the main work and innovations are as follows: firstly, this paper analyzed the CHI algorithm’s flaws, determined that the introduction of new parameters will be the improvement direction of the CHI algorithm, and thus proposed a new algorithm based on variance and coefficient of variation. Secondly, experiment to verify the effectiveness of the new algorithm. In terms of language, the experiment in this paper includes two text classification systems, which were Chinese and English. In terms of classifiers, two classifier algorithms were used, which included the KNN classifier and the Naive Bayes classifier. In terms of data types, two distribution types of data were used: balanced datasets and unbalanced datasets. Finally, experiment and result analysis. This paper has conducted 3 comparative experiments and analyzed the results of each experiment. The experimental results obtained are all significantly improved compared to the results before the improvement.

The wetting behavior and the interfacial characterization of Ti32.8Zr30.2Cu9Ni5.3Be22.7 (denoted as ZT3) bulk metallic glass (BMG) alloy and Ti55 substrate were investigated by the sessile drop method at 1023–1223K in a high vacuum. We found that the system had the final equilibrium contact angle around 40°, 27°, 24°, 20° and 18° at the corresponding experimental temperatures. During the experiment, elements in the substrate diffused upward and formed an interfacial layer at the interface, and then the dendritic phase formed in the cooling process. The thickness of the interfacial layer and the height of the dendrite increased with the increasing experimental temperature. Ti and Zr elements are the main components of the interfacial layer. XRD and EDS results show that the compositions of the interface layer are β-Ti (Zr, Cu, Al), which is different from that of droplet and substrate.

The clustering method of time series is based on the measurement of the whole time series, and the traditional clustering method is used for direct clustering. It was only recently that Keough came up with the concept of Shapelet. He later applied this concept to unsupervised time series clustering. This is a time series clustering method based on the feature of sub-series. This method has good clustering accuracy and stability. However, due to the fact that traversal sequences are all sub-sequences, the original violent search algorithm is very complex and difficult to be applied to a large number of time series data sets. In this paper, on the basis of the original method, you introduce the idea of data retrieval in the data preprocessing stage to quickly screen candidate subsequences. Then we introduce a position-aware hashing algorithm -LSH. The method is used to match the sub-sequences of the pre-processed time series data, and the wide representativeness of the high-quality u-Shapelet is used for selection. Therefore, LSH-us is proposed in this paper to speed up the extraction process of feature sub-sequences. Experimental results show that this algorithm can improve the speed of u-Shapelet time series clustering algorithm and ensure the accuracy of clustering.

The extraction process of cobalt using microemulsion system saponified with P507 as extracting agent was studied. The influences of initial concentration, the concentration of P507, initial pH, the extracting time and the volume ratio of organic phase to water phase on extraction coefficient were also studied. Experimental results show that the saponification microemulsion system using P507 as extractant is thermodynamic stable, and it can separate cobalt better. The optimum conditions were achieved as: the initial concentration is 0.04 mol/L, pH is 2; The concentration of P507 is 6%, extraction time is 6 min; The volume ratio of organic phase to water phase is 1.5:1.

In the light of the shortcomings of traditional wireless data acquisition structure, the thesis analyzes the advantages of data transmission by GPRS, and then a remote monitoring system based on GPRS technology is proposed. The thesis focuses on overall structure of the system, the circuit design and work process of the remote terminal. It describes the realization method of using Socket multithreading technology to achieve network communication in data center.

Biomass char was prepared from mulberry branches by physical activation. An examination by Fourier transform infrared spectroscopy (FTIR) indicated that the functional groups of Si-O were mostly burnt out, significantly decreasing the ash content Analysis of data from a scanning electron microscope (SEM) and a Brunauer-Emmett-Teller (BET) test also revealed increased surface roughness and pore structure, which improved the adsorption capacity of biomass char after preparation. The optimum conditions for preparation were found to be pyrolysis at 700 degrees C for 30 minutes, and then activation at 750 degrees C for one hour, with 3.4% steam content for the activating agent. The prepared biomass char was then employed to adsorb ammonium, copper(II) actetate [Cu(II)] and hexavalent chromium [Cr(VI)] in a solution. The results indicated that the prepared biomass char had a better adsorptive performance than the raw material. Moreover, the removal of determinands increased along with the dosage, and the highest adsorption efficiency of ammonium, copper(II) acetate [Cu(II)] and hexavalent chromium [Cr(VI)] were found to be 20%, 100% and 50%, respectively. The adsorptions of ammonium and hexavalent chromium [Cr(VI)] can be simulated by a pseudo-second order model, while the adsorption of copper(II) acetate [Cu(II)] is better simulated by a pseudo-first order model. The adsorption isotherms of copper(II) acetate [Cu(II)] by biomass char were also investigated, and the Langmuir isotherm was found to best describe the adsorption process.

This paper investigates the resource allocation of Orthogonal Frequency Division Multiplexing Access (OFDMA) cellular network with DF relaying in the downlink. A joint scheme of relay selection, subcarrier assignment and power allocation is proposed to maximize the system sum rate subject to individual power constraints and heterogeneous users' rate requirements. Since the original problem is a mixed integer programming (MIP), we transform it to a standard convex optimization by continuous relaxation and solve it via iterative water-filling (IWF) and subgradient methods in the dual domain. Beside the joint scheme, the upper bound is studied through an exhaustive bisection search. Simulations results show that the proposed scheme outperforms the traditional methods in terms of both system throughput and users' satisfactions.

Rongalite® promotes cleavage of diaryl disulphides generating the corresponding thiolate species in situ which then undergo facile ring-opening of epoxides in a regioselective manner under ultrasound irradiation, affording β-hydroxy sulphides in good to excellent yields. The important features of this methodology are base-free, odourless, high yield, reasonably rapid reaction rate, simple workup, high regioselectivity, cost-effective and no requirement of transition metal catalysts. It is noteworthy that ring-opening reaction of 1,2-diphenyldiselane with 2-(phenoxymethyl)oxirane are also conducted smoothly to afford β-hydroxy selenide in excellent yield under the standard conditions.

This paper discusses the ratio estimator and regression estimator of population mean in the three-stage sampling, in which all level of unit sizes are equal, when some auxiliary information can be available.

This paper concerns the problem of delay-dependent stability criteria for recurrent neural networks with time varying delays. By taking more information of states and activation functions as augmented vectors, a new class of the Lyapunov functional is proposed. Then, some less conservative stability criteria are obtained in terms of linear matrix inequalities (LMIs). Finally, two numerical examples are given to illustrate the effectiveness of the proposed method.

The problem of bounded-input bounded-output (BIBO) stabilization for discrete-time uncertain system with time delayis investigated. By constructing an augmented Lyapunov function, some su cient conditions guaranteeing BIBO stabilizationand robust BIBO stabilization are established. These conditions are expressed in the forms of linear matrixinequalities (LMIs), whose feasibility can be easily checked by using Matlab LMI Toolbox. Two numerical examples areprovided to demonstrate the e ectiveness of the derived results.

This paper specifically develops a framework based on multi-agent automated negotiation. Compared with the popular negotiation model , such as multi-agent cooperation negotiation model reported by Sycara[2], agent automated negotiation model by Paurobally[3], and negotiation model based on deadline by Wooldridge[4], the main idea in this model is that negotiation is of multi-factor , negotiation agent inherits from multi-agent ,and negotiation for e-commerce model is multiagent with multi-factor. During the deal , most agents such as buyers, sellers and intermediates take part in negotiation. Agent acts differently in different negotiations , for example , during English auction, auctioneer acts as a dominant factor. Every agent has to calculate its own preferences , which are described as role , rationality , knowledge , commitment , social behavior and bidding strategy, to select from different deals[5]. This paper will discuss the optimal negotiation strategies in a certain negotiation case and three-dimensional space.

The robust stability of uncertain discrete-time recurrent neural networks with time-varying delay is investigated. By decomposing some connection weight matrices, new Lyapunov-Krasovskii functionals are constructed, and serial new improved stability criteria are derived. These criteria are formulated in the forms of linear matrix inequalities (LMIs). Compared with some previous results, the new results are less conservative. Three numerical examples are provided to demonstrate the less conservatism and effectiveness of the proposed method.

In this paper, a class of linear switched systems with time delays is investigated. At the same time, the impulsive effects and parametric uncertainty assumed to be norm-bounded are considered. Some novel exponentially stable criteria are obtained by constructing appropriate Lyapunov-Krasovskii functional. The validity of the results is shown by one numerical example.

In view of the process of integration of e-government information resources across heterogeneous platforms chief departments of information sharing and information exchange, this paper proposed a method of e-government information resources integration based on the directory service. It presents a technical framework of the directory service system which includes resource layer and directory layer and service layer and application layer. It gives the role model of the process of realizing the directory service based on UDDI technology, and discussed the technology plan and exchange model of e-government information exchange system. It analyzes the features of the exchange patterns which are centralism exchange pattern and distribution exchange pattern and mixture pattern.

Constructing an effective translation between different representations is the key point to overcome semantic heterogeneity in information integration systems. This paper puts forward an ontology based information representation architecture-SOSRA. It introduces a GAV (global as view) approach to solve the mapping between heterogeneous semi-structured data in the form of service, discusses the semantic representation algorithm and semantic mapping strategy emphatically. The implementation of SOSRA and its application to semantic integration system is given at the end of this paper to enhance information process accuracy and efficiency.