基于1D CNN与XGBoost的恶意代码纹理检测

恶意代码数量已经呈现爆炸式增长,对于恶意代码的检测防护显得尤为重要.近几年,基于深度学习的恶意代码检测方法开始出现,基于此,提出一种新的检测方法,将恶意代码二进制文件转化为十进制数组,并利用一维卷积神经网络(1 Dimention Convolutional Neural Networks,1D CNN)对数组进行分类和识别.针对代码家族之间数量不平衡的现象,该算法选择在分类预测上表现良好的XGBoost,并对Vision Research Lab中的25个不同恶意软件家族的9458个恶意软件样本进行了实验.实验结果表明,所提的方法分类预测精度达到了97％.     

Dense and core-rim structured B4C-TiB2 ceramics with Mo-Co-WC additive

B₄C-TiB₂ ceramics (TiB₂ ranging 5~70 vol%) with Mo-Co-WC as the sintering additive were prepared by spark plasma sintering. In comparison with B₄C-TiB₂ without additive, the enhanced densification was evident in the sintered specimen with Mo-Co-WC additive. Core-rim structured grain was observed around TiB₂ grains. The interface of the rim between TiB₂ and B₄C phases demonstrated different feature: the inner borderline of the rim exhibited a smooth feature, whereas a sharp curved grain boundary was observed between the rim and the B₄C grain. The formation mechanism is discussed: the epitaxial growth of (Ti,Mo,W)B₂ rim around the TiB₂ core may occur as a result of the solid solution and dissolution-precipitation between TiB₂ phase and the sintering additive. It was revealed that the fracture toughness increased as the content of TiB₂ content increased, alongside the decreased hardness. B₄C-30 vol% TiB₂ specimen demonstrated the optimal combination of mechanical properties, reaching Vickers hardness of 24.3 GPa and fracture toughness of 3.33 MPa·m^1/2.     

Mechanical Properties of Single-Layer Diamond Reinforced Poly(vinyl alcohol) Nanocomposites through Atomistic Simulation

Low-dimensional carbon nanostructures are ideal nanofillers to reinforce the mechanical performance of polymer nanocomposites due to their excellent mechanical properties. Through molecular dynamics simulations, the mechanical performance of poly(vinyl alchohol) (PVA) nanocomposites reinforced with a single-layer diamond – diamane is investigated. It is found the PVA/diamane exhibits similar interfacial strengths and pull-out characteristics with the PVA/bilayer-graphene counterpart. Specifically, when the nanofiller is fully embedded in the nanocomposite, it is unable to deform simultaneously with the PVA matrix due to the weak interfacial load transfer efficiency, thus the enhancement effect is not significant. In comparison, diamane can effectively promote the tensile properties of the nanocomposite when it has a laminated structure as it deforms simultaneously with the matrix. With this configuration, the interlayer sp³ bonds endows diamane with a much higher resistance under compression and shear tests, thus the nanocomposite can reach very high compressive and shear stress. Overall, enhancement on the mechanical interlocking at the interface as triggered by surface functionalization is only effective for the fully embedded nanofiller. This work provides a fundamental understanding of the mechanical properties of PVA nanocomposites reinforced by diamane, which can shed lights on the design and preparation of next generation high-performance nanocomposites.     

A practical reliability assessment approach and its application for pile-stabilized slopes using FORM and support vector machine

Bulletin of Engineering Geology and the Environment - Many uncertainties exist in pile-stabilized slopes which make their design substantially complicated. In this paper, a first-order reliability...     

Software assurance practices for mobile applications

Mobile software applications have to cope with a particular environment that involves small size, limited resources, high autonomy requirements, competitive business models and many other challenges. To provide development guidelines that respond to these needs, several practices have been introduced; however, it is not clear how these guidelines may contribute to solve the issues present in the mobile domain. Furthermore, the rapid evolution of the mobile ecosystem challenges many of the premises upon which the proposed practices were designed. In this paper, we present a survey of the literature on software assurance practices for mobile applications, with the objective of describing them and assessing their contribution and success. We identified, organized and reviewed a body of research that spans in three levels: software development processes, software product assurance practices, and software implementation practices. By carrying out this literature survey, we reviewed the different approaches that researchers on Software Engineering have provided to address the needs that raise in the mobile software development arena. Moreover, we review the evolution of these practices, identifying how the constant changes and modernization of the mobile execution environment has impacted the methods proposed in the literature. Finally, we introduced discussion on the application of these practices in a real productive setting, opening an area for further research that may determine if practitioners have followed the proposed assurance paradigms.     

Intrinsic CFRP-metal-hybrids with rubber interface for the improvement of the damping behaviour

This paper presents an integrated passive damping approach in hybrid metal-CFRP parts for structural applications. In this concept a viscoelastic material is embedded in the joint zone of the hybrid component. To examine the connection strength single-lap-joint specimens were produced and tested and the influence of the used material combinations, different surface structures, and different process parameters i.e. the moment of cross-linking were evaluated. Afterwards, the metal-CFRP hybrids were tested in quasi-static tests to assess their connection strength and failure behaviour. Dynamic cyclic tensile tests with step-wise increased loading conditions were performed to determine the specimens damping behaviour and to estimate their fatigue performance. Finally, these results are compared to a state of the art metal-CFRP hybrid with rivets connecting both materials.     

On stochastic comparisons of finite mixtures for some semiparametric families of distributions

In this paper, we consider the classical finite mixture model, which is an effective tool for modeling lifetime distributions for random samples from heterogeneous populations. We discuss new results on stochastic comparison for two finite mixtures when each of them is drawn from one of the following semiparametric families, i.e., proportional hazards, accelerated lifetime and proportional reversed hazards.