Manifold elastic net: a unified framework for sparse dimension reduction

It is difficult to find the optimal sparse solution of a manifold learning based dimensionality reduction algorithm. The lasso or the elastic net penalized manifold learning based dimensionality reduction is not directly a lasso penalized least square problem and thus the least angle regression (LARS) (Efron et al., Ann Stat 32(2):407–499, 2004), one of the most popular algorithms in sparse learning, cannot be applied. Therefore, most current approaches take indirect ways or have strict settings, which can be inconvenient for applications. In this paper, we proposed the manifold elastic net or MEN for short. MEN incorporates the merits of both the manifold learning based dimensionality reduction and the sparse learning based dimensionality reduction. By using a series of equivalent transformations, we show MEN is equivalent to the lasso penalized least square problem and thus LARS is adopted to obtain the optimal sparse solution of MEN. In particular, MEN has the following advantages for subsequent classification: (1) the local geometry of samples is well preserved for low dimensional data representation, (2) both the margin maximization and the classification error minimization are considered for sparse projection calculation, (3) the projection matrix of MEN improves the parsimony in computation, (4) the elastic net penalty reduces the over-fitting problem, and (5) the projection matrix of MEN can be interpreted psychologically and physiologically. Experimental evidence on face recognition over various popular datasets suggests that MEN is superior to top level dimensionality reduction algorithms.     

Enhanced opposition-based differential evolution for solving high-dimensional continuous optimization problems

This paper presents a novel algorithm based on generalized opposition-based learning (GOBL) to improve the performance of differential evolution (DE) to solve high-dimensional optimization problems efficiently. The proposed approach, namely GODE, employs similar schemes of opposition-based DE (ODE) for opposition-based population initialization and generation jumping with GOBL. Experiments are conducted to verify the performance of GODE on 19 high-dimensional problems with D = 50, 100, 200, 500, 1,000. The results confirm that GODE outperforms classical DE, real-coded CHC (crossgenerational elitist selection, heterogeneous recombination, and cataclysmic mutation) and G-CMA-ES (restart covariant matrix evolutionary strategy) on the majority of test problems.     

Next generation session management for 3D teleimmersive interactive environments

In the recent past we have seen a boom of distributed interactive multimedia environments which use multiple correlated media sensors, multi-view displays, and advanced haptic-audio-visual user interfaces for teleimmersive gaming, business meetings and other collaborative activities. However, when we investigate the emerging teleimmersive environments closer, we realize that their overall session management, including end-to-end session setup, establishment and run-time management are not appropriate for the new demands that these environments present. These environments are cyber-physical rooms that demand (a) large scale of multi-sensory devices across geographically-distributed locations and (b) interaction with each other in synchronous and real-time manner. To deal with the new teleimmersive demands, we present a new session management design with (a) session initiation protocol(s) that understand media correlations, (b) view-based multi-stream topology establishment among multiple parties, (c) efficient, light-weight and distributed session monitoring with querying and debugging capabilities, (d) flexible view-based session adaptation with efficient topology adjustments, and (e) light-weighted and consistent session tear-down protocols. The presented design of the next generation session management protocols, services, algorithms and data structures is based on our extensive experiences with building 3D teleimmersive interactive systems, experimenting with high impact teleimmersive applications and deploying such environments at various venues.     

Membership evaluation and feature selection for fuzzy support vector machine based on fuzzy rough sets

A fuzzy support vector machine (FSVM) is an improvement in SVMs for dealing with data sets with outliers. In FSVM, a key step is to compute the membership for every training sample. Existing approaches of computing the membership of a sample are motivated by the existence of outliers in data sets and do not take account of the inconsistency between conditional attributes and decision classes. However, this kind of inconsistency can affect membership for every sample and has been considered in fuzzy rough set theory. In this paper, we develop a new method to compute membership for FSVMs by using a Gaussian kernel-based fuzzy rough set. Furthermore, we employ a technique of attribute reduction using Gaussian kernel-based fuzzy rough sets to perform feature selection for FSVMs. Based on these discussions we combine the FSVMs and fuzzy rough sets methods together. The experimental results show that the proposed approaches are feasible and effective.     

Research on the group decision-making about emergency event based on network technology

In order to improve decision-making efficiency about emergency event, this paper proposes a novel concept, i.e., Agile-Delphi Method, which is an integration of agile decision and Delphi Method implicating that the decision-makers instantly deliver, respond, treat, and utilize information via Delphi process while conducting group decision-making about emergency event. The paper details the mechanism of group decision-making about emergency event based on network technology and Agile-Delphi Method. Finally, the paper conducts an empiric analysis taking the “111 event”, i.e., the liquid ammonia spill event happened on November 1, 2006 in a phosphorus chemical company in China, as an example.     

Simulation of solid thermal explosion and liquid thermal explosion of dicumyl peroxide using calorimetric technique

Dicumyl peroxide (DCPO), is produced by cumene hydroperoxide (CHP) process, is utilized as an initiator for polymerization, a prevailing source of free radicals, a hardener, and a linking agent. DCPO has caused several thermal explosion and runaway reaction accidents in reaction and storage zone in Taiwan because of its unstable reactive property. Differential scanning calorimetry (DSC) was used to determine thermokinetic parameters including 700 J g^–1 of heat of decomposition (ΔH_d), 110 °C of exothermic onset temperature (T₀), 130 kJ mol^–1 of activation energy (E_a), etc., and to analyze the runaway behavior of DCPO in a reaction and storage zone. To evaluate thermal explosion of DCPO with storage equipment, solid thermal explosion (STE) and liquid thermal explosion (LTE) of thermal safety software (TSS) were applied to simulate storage tank under various environmental temperatures (T_e). T_e exceeding the T₀ of DCPO can be discovered as a liquid thermal explosion situation. DCPO was stored under room temperature without sunshine and was prohibited exceeding 67 °C of self-accelerating decomposition temperature (SADT) for a tank (radius = 1 m and height = 2 m). SADT of DCPO in a box (width, length and height = 1 m, respectively) was determined to be 60 °C. The TSS was employed to simulate the fundamental thermal explosion behavior in a large tank or a drum. Results from curve fitting demonstrated that, even at the earlier stage of the reaction in the experiments, ambient temperature could elicit exothermic reactions of DCPO. To curtail the extent of the risk, relevant hazard information is quite significant and must be provided in the manufacturing process.     

Implementation of a transient adaptive sub-cell module for the parallel-DSMC code using unstructured grids

A method of transient adaptive sub-cells (TAS) suitable for unstructured grids that is modified from the existing one for the structured grids of DSMC is introduced. The TAS algorithm is implemented within the framework of a parallelized DSMC code (PDSC). Benchmarking tests are conducted for steady driven cavity flow, steady hypersonic flow over a two-dimensional cylinder, steady hypersonic flow over a cylinder/flare and the unsteady vortex shedding behind a two-dimensional cylinder. The use of TAS enables a reduction in the computational expense of the simulation since larger sampling cells and less simulation particles can be employed. Furthermore, the collision quality of the simulation is maintained or improved and the preservation of property gradients and vorticity at the scale of the sub-cells enables correct unsteady vortex shedding frequencies to be predicted. The use of TAS in a parallel-DSMC code allows simulations of unsteady processes at a level to be carried out efficiently, accurately and with acceptable computational time.     

实时刚体破碎特效仿真研究

刚体仿真是图形学研究的热点之一,但是单纯依据物理规律进行仿真达不到实时性要求.依据攻击武器的运动轨迹在目标的八叉树结构上抽取空间连续的子块,设计了一种摧毁区域锁定算法;提出一种刚体预破碎模式,采用线性约束连接子块和块内元素,并用统一的算法检测由攻击武器的碰撞引起的约束动态受损情况.线性约束的不同程度破损或导致子块的分裂,或导致块内基础元素的分裂,将形成不同程度的破碎特效.最后设计了实时刚体破碎特效实验方案并构建仿真系统.实验表明,该种模式不仅较好地体现了刚体破碎特效,也简化了刚体破碎仿真的复杂性,相应的破碎特效能够拓展到具有实时性需求的虚拟现实应用环境中.     

BeehiveZ:一个开放的业务过程模型管理框架

随着面向服务架构、业务过程管理技术的推广应用,企业中用建模语言描述的业务过程模型越来越多,这些模型描述了企业核心的业务逻辑,是企业最宝贵的知识资产.为了更好地促进企业的发展,提升竞争力,需要将这些业务过程模型作为数据进行有效地管理、建立企业过程模型库,并支持对过程模型的存储、浏览、相似性度量、检索等操作和应用开发.提出了一个开放的业务过程模型管理框架--BeeHiveZ,阐述了BeeHiveZ的体系结构、关键技术及系统功能与操作等内容.BeeHiveZ现已在开源软件社区SourceForge. net上发布.     

InfoSigs:一种面向Web对象的细粒度聚类算法

面向Web对象的细粒度聚类已经成为学术界研究的热点.然而现有大多数聚类模型只关注如何对文本内容或文章主题进行聚类,聚类结果粒度较粗,无法满足大规模网络信息检索的质量要求.针对上述挑战,充分挖掘Web文档中词汇间的树状概率层次关系,提出一种以词汇信息分布作为特征标志的聚类算法InfoSigs,实现对Web对象的细粒度聚类.算法构建一个信息传递有向无环图,根据词汇在图中信息分布的集中度赋予其合理的权重,产生更具代表性的特征向量;同时算法提出了一个自适应的记录合并模型,有效提高记录簇中记录间的相似度,减少噪音对合并过程的影响.实验结果表明,InfoSigs算法比传统聚类算法—I-Match和Shingling—在F-Measure值上平均约有21.3%的提高,可以有效地运用到多领域Web对象的聚类问题.