利用最小冗余最大相关和SVM的SAR图像海上溢油识别

近年溢油事故频发,海洋污染日益严重。利用合成孔径雷达（SAR）卫星可以有效跟踪由海上溢油事故导致的油膜扩展情况。利用最大冗余最小相关-支持向量机（mRMR_SVM）算法进行SAR图像溢油识别,为溢油事故决策支持提供重要前提。首先采用mRMR提取最优特征向量集,对输入值进行降维; 然后采用SVM算法解决油膜图像分类问题,同时选择径向基函数（RBF）为核函数;使用训练集训练该模型,调整模型参数;以测试集特征向量作为输入,利用训练好的模型进行溢油识别。实验结果表明,mRMR_SVM模型对SAR图像的油膜和类油膜识别有效,准确率为96.875%.     

基于改进指纹聚类的WLAN定位优化方法

将K-means聚类算法应用到无线局域网（WLAN）位置指纹定位中，虽然可以缩短定位时间，但是容易降低定位精度。为了解决此问题，提出了基于改进指纹聚类的WLAN定位优化方法。首先根据接收信号强度标准差来优化初始聚类中心的选取,然后对指纹数据进行聚类处理,最后进行在线定位。实验结果表明，与传统的WLAN位置指纹定位方法和K-means聚类定位方法相比，基于改进指纹聚类的定位优化方法不仅缩短了定位时间，还能有效提高定位精度。     

一种基于神经网络提取旋波材料参数的新方法

提出了一种从测试的S参数中提取旋波材料参数的新方法——神经网络法。针对旋波材料参数的自由空间测试方法，本文采用BP神经网络来构建映射模型，利用不同算法对网络进行训练，从而获得所需的网络。结果表明，该方法计算速度快、精度高，解决了求解非线性方程组解的模糊性问题，确保了测试结果的可靠性。该方法所构建的网络可用于材料测试后旋波材料参数的快速提取。     

一种新的软聚类方法在入侵检测中的应用

本文在FC(MFuzzyC-Means)算法的基础上结合自适应和模糊截集理论提出了一种新的算法S2AFC(MSectionSetAdaptiveFCM),以及这个算法针对进程轨迹序列在异常检测中的应用。本算法能够克服聚类数目c要求预先确定、局部最优、分类不确定等弱点,在聚类过程中可动态调整并自动获得最佳的聚类数目。最后用实验和相关数据证明了本算法的正确性和有效性。结果表明本文所提出的算法能够有效地检测入侵,并在性能上有所提高。     

基于改进K-means聚类和量子粒子群算法的多航迹规划

针对在复杂环境下需要通过多航迹规划以实现武器协同的问题,利用排挤机制产生K-means聚类的初始聚类中心,并将改进K-means聚类与量子粒子群算法（QPSO）相结合应用于无人机的三维多航迹规划。改进算法解决了K-means聚类易陷入局部最优、聚类准确率低的问题。根据产生的初始聚类中心,将粒子划分成多个子种群,利用QPSO算法对每个子种群进行优化,使得每个子种群可以产生一条可行航迹。仿真分析证明了改进算法可以有效保证子种群之间的多样性,生成较为分散的多条可行航迹。     

BP算法在股指期货合约价格预测中的应用

本文将Matlab软件中BP算法工具箱中的newff函数引入到了股指期货合约价格的预测分析中,利用BP网络的数据压缩性以及非线性映射功能,模拟股指期货合约结算价与最高价、最低价、开盘价、收盘价、总持仓量、真实指数值、期货合约成交总金额之间的关系,建立了基于神经网络的股指期货合约结算价预测系统。用交易数据进行网络训练,生成网络后,用另外的数据对生成的网络进行验证,得出了该网络的预测精度较高,具备一定的应用价值。     

微粒群算法在瓦斯突出预测中的应用

如何确定RBF神经网络的隐单元RBF的中心c1直接影响到神经网络的逼近能力。本文通过带惯性权重的微粒群算法和学习训练法相结合的方式来确定RBF神经网络的隐单元RBF的中心c1。首先使用学习训练的方法根据局部信息确定中心c1,然后充分发挥了微粒群算法全局寻优的特点,通过时数据库的更新,使用带惯性权重的微粒群算法对得到的中心c1进行最优化．同时使用MATLAB进行程序设计。最后将RBF神经网络用于瓦斯突出预测中,得出了准确的预测结果。     

采用局部二值模式与深度信念网络的人脸识别

针对人脸维度过高和人脸局部特征提取易忽略的问题,提出了一种将多尺度局部二值模式（LBP）算法与深度信念网络(DBN)算法相结合的人脸识别方法。首先采用多尺度LBP算法提取人脸纹理特征,进而将LBP提取的纹理特征作为深度信念网络的输入,最后通过逐层网络训练,得到网络的最优参数,并在ORL人脸库中进行测试,识别率可达95.2%,比使用Gabor小波和主成分分析（PCA）算法的人脸识别高2.6%,说明该算法具有很好的人脸识别能力。     

基于MEA-GRNN的RFID室内定位

针对采用遗传算法(Genetic Algorithm,GA)优化广义回归神经网络(Generalized Regression Neural Network,GRNN)的射频识别（Radio Frequency Identification，RFID）室内定位模型存在的早熟、收敛速度慢、不能保证解是全局最优等问题，提出采用思维进化算法(Mind Evolutionary Algorithm,MEA)来寻找广义神经网络的最优光滑因子，从而确定最优定位模型。首先用GRNN建立节点定位模型，阅读器与标签的接收信号强度值作为GRNN的输入，节点坐标作为输出，根据适应度函数值，通过MEA寻找GRNN的最优平滑参数。实验结果表明，通过MEA优化的GRNN模型的定位精度比GA优化的GRNN定位模型的精度高、泛化能力强，并且比后者的效率高,能够避免GA陷入局部最优的问题。     

基于PSO-BP算法的无线传感器网络定位优化

在研究现有定位算法的基础上，针对基于接收信号强度指示(RSSI)定位模型中的参数易受环境影响等问题，提出了一种新型的粒子群优化(PSO)算法与后向传播（BP）神经网络相结合的算法。BP网络算法权值的修正依赖于非线性梯度值，易形成局部极值，同时学习次数较多，需先通过粒子群算法进行优化。为了提高定位精度，首先采用速度常量法滤波处理，然后通过改进的混合优化算法对BP神经网络初始权值和阈值进行优化，并分析算法的性能。试验中隐层节点个数采用试错法，从12到19变化，以确定合适数目。实验结果表明，与一般加权算法和传统BP算法相比，改进的混合优化算法可大幅改善测距误差对定位误差的影响，同时可使25 m内最小定位误差小于0.27m     

Brown＇s Blossom in Britain-China

The overall strategic relationship between U.K. and China is ＂better than ever＂, China- Britain relations have maintained good momentum, with increased high-level exchanges and the reinforcement of strategic mutual trust, with more room for growth in the future.     

Challenge and Opportunity Go Hand in Hand——What Global Economic Agitation Brings to the China's Economy?

Domino effects of the global finance agitation As the origin of the subprime crisis,US financial institutions lost USS45 billion last year,accounted for over half of the total US,EU & Japan loss.Since January 15th,Chtigroup released its financial statement of the 4th quarter,showing a net deficit of USS9.83 billion,which marked the hisoric high of its quarterly deficit.     

Sinosteel：Extraordinary Moves in Industry Chess

Among the Fortune 500 firms,more and more images from China emerge out.On the 2008 list,Sinosteel Corporation,a landmark group of China industries quietly squeezes into the candidate line.As the largest steel service ＆ trade enterprise,its sales reached at RMB 111.24 billion,with a year-on-year increase of 83.03 percent,and the profit grew by 180.03 percent.On the Sinosteel 2008 Work Meeting,the president Huang Tianwen uttered,＂In the coming year,we will work for US$20 billion sales,and look forward to getting a position in Fortune 500 firms for ourselves.＂     

Walk Through Unusual Five Years—— An Exhibitor＇s Canton Fair Course

When the 103rd Canton Fair is approaching, an old friend of Canton Fair, Li Ying from Jiangsu prov-ince recollected his past five years＇ experiences to the grand trade meeting. Each year is special, with some unforgettable memories colored by the year.     

Universal Coverage, Pinpoint Illustration and Profound Impact HTC 2008 Intermedia Recommendation and Introduction Meeting

Holding 88 percent of train media resources across China, based on "Time Cube", on the 2008 Intermedia Recommendation and Introduction Meeting on December 19, 2007, HTC announced its high-profile strategy of tour marketing and tapping the immeasurable potentials on the train. Meanwhile, HTC officially launched its brand new product Olympic Train on the meeting, providing a fabulous platform for all the brands that are prepared to do a smart marketing on the occasion of 2008 Olympic Games.     

Market presence in China strengthened Network growth and Dual-Chinese-Hub Strategy More frequet Flyer Benefit——Air China and shanhai Airlines Join Star Alliance

In the end of 2007, Air China Limited (hereinafter "Air China") and Shanghai Airlines, formally joined Star Alliance at a ceremony held in the new Terminal Three at Beijing Capital International Airport. With the addition of the two Chinese airlines, Star Alliance now has 19 member carriers operating 17,000 daily flights to 897 destinations in 160 countries.     

China''''s Stock Market Wave

The annual meeting of the World Economy Forum was held in Davos,a little town in Switzerland,on January 23rd to the 27th this year,with the theme"the Power fo Collaborative Innovation",over 2500 celebrates from ture political and cconomic circles of more than 80 countries and challengred logether to talk about the future of the global economy in 2008,A Forum for the problems and challenges that face whole world,some of the keywords which drew attention during the event were:the U.S.subprime mortgage crisis,soaring energy prices,the economic slowdown in the U.S.ermerging economic bodies,authority funds and the general unpredictable nature of the global economy.     

Trade Promotion from Neighborhood

If you are considering to trade with companies in Japan or expanding your .business to Japan, we are ready to help from the very beginning.＂ You will find this greeting at the first sight on the Chinese website of Japan External Trade Organization （JETRO）. JETRO has committed itself in delivering highly efficient and quality services to meet the demand of internal and external customs including Japanese companies, hence winning a good reputation in the international trade arena. In recent years, with its priority in advancing business cooperation in South and East Asia, JETRO has made great achievement in promoting trade between China and Japan. Therefore, China＇s Foreign Trade interviewed Mr. Yoichi Maine, Deputy Director General of JETRO BEIJING.     

Auto Parts Market Potential Coming into Surface

Looking at the increasingly fierce competition in the world auto market. global automobile manufacturing industry, taking Europe, the U.S. and Japan as representatives, is gradually forming a new type of relationship, mutual cooperation between whole vehicle and parts manufacturers. We can clearly see that auto parts manufacturers have become independent from the whole vehicle manufacturing companies, which means that the former structure of auto industry has greatly changed, and a new relationship,     

Jean-Pierre Duprieu： Willing to Be A Legate of Sino-French Business

Estaing, former French president, Valery Giscard d＇Estaing attended Sino-French Economic Seminar held in Shanghai in early April and said that the world center has shifted from the west to the east with many European and US companies entering into Asia, and China has the ability to play an important role in the world stage. Estaing, naming himself＂French Confucian＂, then said in Chinese that a cook-group includes 10 people in ancient Chinese soldier system and the cook is in charge of cooking for them. And people who belong to the same cook are actually ＂Partners＂.