基于自组织神经网络的实时统计特征抽取方法

应用人工神经网络讨论统计了特征抽取的原理，提出了相应的抽取方法。该方法具有实时性，容错性，高数据压缩率和良好的特征分辨率。通过计算机仿真和桥梁索力状态的实例计算分析证实了该技术具有广泛的应用价值。     

Synthesis and characterization of block copolymers from D,L-lactide and poly(tetramethylene ether glycol)

The copolymer from D ,L -lactide and poly(tetramethyene ether glycol) (PTMG) was prepared in bulk with an isotributyl aluminum–water–phosphoric acid complex catalyst as the initiator and characterized by H-NMR, GPC, and DSC. The effects of the temperature and the amount of PTMG on the polymerization rate and the molecular weight of copolymers were studied. The behavior of the degradation and delivery rate of Levonorgestrel microspheres in vitro was observed. The results show that the degradation and the delivery rate can be controlled by adjusting the molar rate of hydrophilic and hydrophobic segments of the copolymer © 1995 John Wiley & Sons, Inc.     

Chemical constituents of Luffa cylindrica (L.) Roem

Six compounds were isolated from the fruits of Luffa cylindrica. They were identified as lucyosides C, E, F, H, a mixture of alpha-spinasterol and stigmasta-7,22,25-trien-3 beta-OH and a mixture of alpha-spinasteryl glucoside and delta 7,22,25-stigmasteryl-beta-D-glucoside by means of chemical evidence and spectral analysis.     

北京城区河流沉水植物分布特征及环境因子关系研究

通过北京城区11条河流30个点位沉水植物及水环境因子调查,对沉水物物种分布、生态位特征及与水环境因子相互关系进行了研究,结果表明:北京城区河流均有沉水植物分布,春季优势种为菹草、穗花狐尾藻和篦齿眼子菜,夏季为金鱼藻、狐尾藻、篦齿眼子菜、黑藻、苦草及菹草.水深为影响城区河流沉水植物种类数量的重要因子,菹草为春季沉水植物管理关键物种.研究城区河流沉水植物分布特征及与水环境因子的关系,对北京市河流水生态环境的管理和持续改善具有重要的参考和指导意义.     

Entity-level sentiment prediction in Danmaku video interaction

The Journal of Supercomputing - Sentiment analysis in Danmaku video interaction aims at measuring public mood in respect of the video, which is helpful for the potential applications in behavioral...     

A frequency-aware and energy-saving strategy based on DVFS for Spark

The Journal of Supercomputing - With the fast growth of big data applications, it has brought about a huge increase in the energy consumption for big data processing in Cloud data centers. In this...     

Evaluating Visual Properties via Robust HodgeRank

International Journal of Computer Vision - Nowadays, how to effectively evaluate visual properties has become a popular topic for fine-grained visual comprehension. In this paper we study the...     

A low distortion image enhancement scheme based on multi-resolutions analysis in next generation network

With the advent of Next Generation Network (NGN), services that are currently provided by multiple specific network-centric architectures. NGN provides AAA (Anytime, Anywhere and Always on) access to users from different service providers with consistent and ubiquitous provision of services as necessary. This special issue of NGN includes pervasive, grid, and peer-to-peer computing to provide computing and communication services at anytime and anywhere. In fact, the application of NGN includes digital image processing, multimedia systems/services, and so on. Here we focus on the digital image processing technology in NGN environments. Low-contrast structure and heavy noise in NGN environments can be found in many kinds of digital images, which makes the images vague and uncertainly, especially in x-ray images. As result, some useful tiny characteristic are weakened—which are difficult to distinguish even by naked eyes. Based on the combination of no-linear grad-contrast operator and multi-resolution wavelet analysis, a kind of image enhancement processing algorithm for useful tiny characters is presented. The algorithm can enhance the tiny characters while confine amplifying noise. The analysis of the results shows that local regions of the image are enhanced by using the concept of the grad contrast to make image clearer adaptively. Experiments were conducted on real pictures, and the results show that the algorithm is flexible and convenient.     

Mining blackhole and volcano patterns in directed graphs: a general approach

Given a directed graph, the problem of blackhole mining is to identify groups of nodes, called blackhole patterns, in a way such that the average in-weight of this group is significantly larger than the average out-weight of the same group. The problem of finding volcano patterns is a dual problem of mining blackhole patterns. Therefore, we focus on discovering the blackhole patterns. Indeed, in this article, we develop a generalized blackhole mining framework. Specifically, we first design two pruning schemes for reducing the computational cost by reducing both the number of candidate patterns and the average computation cost for each candidate pattern. The first pruning scheme is to exploit the concept of combination dominance to reduce the exponential growth search space. Based on this pruning approach, we develop the gBlackhole algorithm. Instead, the second pruning scheme is an approximate approach, named approxBlackhole, which can strike a balance between the efficiency and the completeness of blackhole mining. Finally, experimental results on real-world data show that the performance of approxBlackhole can be several orders of magnitude faster than gBlackhole, and both of them have huge computational advantages over the brute-force approach. Also, we show that the blackhole mining algorithm can be used to capture some suspicious financial fraud patterns.