联合开发网   搜索   要求与建议
                登陆    注册
排序按匹配   按投票   按下载次数   按上传日期
按平台查找All PDF(12) 

[人工智能/神经网络/深度学习] Optimization and KKT conditions

由复旦大学经济学院中国经济研究中心冯曲编写的最优化与KKT条件一书
The best book of optimization and KKT conditions written by Feng Qu, the center of Economic Research of Fudan University, China Economic Research Center (2018-01-17, PDF, 236KB, 下载11次)

http://www.pudn.com/Download/item/id/1516172404864187.html

[人工智能/神经网络/深度学习] Clustering-Algorithms

在众多聚类算法中,基于划分的模糊聚类算法是模式识别中最常用的算法类型之一.至今,文献中仍不断 有关于基于划分的模糊聚类算法的研究成果出现.为了能更为系统和深入地了解这些聚类算法及其性质,本文从改 变度量方式、改变约束条件、在目标函数中引入熵以及考虑对聚类中心进行约束等几个方面,对在 C-均值算法的 基础上得到的基于划分的模糊聚类算法作了综述和评价,对各典型算法的优缺点进行了实验比较分析.指出标准 FCM算法被广泛应用的原因之一是它对数据的比例变化具有鲁棒性,而其他类似的算法对这种比例变化却很敏感, 并以极大熵方法为例进行了比较实验.最后总结了基于划分的模糊聚类算法普遍存在的问题及其发展前景.
Fuzzy partitional clustering algorithms are widely used in pattern recognition field. Until now, more and more research results on them have been developed in the literature. In order to study these algorithms systematically and deeply, they are reviewed in this paper based on c-means algorithm, from metrics, entropy, and constraints on membership function or cluster centers. Moreover, the advantages and disadvantages of the typical fuzzy partitional algorithms are discussed. It is pointed out that the standard FCM algorithm is robust to the scaling transformation of dataset, while others are sensitive to such transformation. Such conclusion is experimentally verified when implementing the standard FCM and the maximum entropy clustering algorithm. Finally, the problems existing in these algorithms and the prospects of the fuzzy partitional algorithms are discussed. (2014-04-03, PDF, 442KB, 下载7次)

http://www.pudn.com/Download/item/id/2501147.html

[人工智能/神经网络/深度学习] gf

白化处理的自然梯度盲源分离统一算法_刘鑫
Natural gradient blind source separation whitening unified algorithm _ Liu Xin (2014-01-13, PDF, 346KB, 下载9次)

http://www.pudn.com/Download/item/id/2449424.html

[人工智能/神经网络/深度学习] TSP110707

本文首先提出了基于TSP 的多点位加工路径优化模型以最高生产效率为目标 函数并以数控中心的多孔位加工和多点位焊接两个典型的自动化加工系统为例来 验证该模型的经济性和适用性其次分别为这两个系统建立了基于TSP 的路径优 化模型模型中考虑了实际加工时的系统中约束条件的影响如数控加工中心的换 刀速度主轴移动速度和焊接机器人的工作台转动速度等提出了近似法将数 控加工中心主轴对同一个孔位n 次重复遍历如钻粗镗精镗等简化对n 个 极小间距的相邻点位的一次遍历从而建立相应的多点位数控加工模型然后分 析和比较了罚函数法Tabu 算法等算法这些算法在求解TSP 时存在收敛速度慢 运算过程复杂的不足本文结合了这些算法的一些特点采用了改进的Hopfield 算 法求解基于TSP 的路径优化模型用C 语言编制了该改进算法的优化程序经实例 试算表明该算法具有收敛快权系数变化小迭代次数少的特点
Firstly, multi-point manufacture routine optimum model is referred. The highest efficient is the object function. Multi-holes machining and multi-weld system are adapted to verify the economics and adaptability of model. Secondly, separately the routine optimum models of both systems are built up. Restriction in the real systems is considered (such as the tool change speed, staple shaft moving speed in CNC and operation platform rotation speed in weld robot). The approximate way is referred to predigest the n times visits as 1 times visit, to establish the multi-holes CNC machining model. Thirdly, analyze and compare the characters and difference of penalty function, Tabu algorithms, point out their faults combine with the merits of other algorithms, and adapt the improving Hopfield algorithms calculate the routine optimum model based on TSP, which is speedily and easily calculated. The optimum program is compiled in C language to achieve the improving algorithms. (2011-07-07, PDF, 511KB, 下载11次)

http://www.pudn.com/Download/item/id/1591799.html

[人工智能/神经网络/深度学习] AlgorithmforWheelRailContactGeometry

提出并实现一种轮轨接触几何算法,可以检测铁道车辆系统动力学仿真在线计算时轮对与钢轨的刚性 单接触斑、多接触斑和跳离情形. 所得刚性接触斑可以为Hertz 理论提供刚性穿透量和曲率,为非Hertz 理论提 供接触区域的法向间隙,为动力学仿真提供接触中心位置和法向方向. 轮对与钢轨的计算机三维图形显示表明 该算法是有效的. 该算法已用于开发空间耦合的铁道车辆系统动力学仿真软件
Algorithm for Wheel Rail Contact Geometry (2010-09-15, PDF, 263KB, 下载85次)

http://www.pudn.com/Download/item/id/1296652.html

[人工智能/神经网络/深度学习] Cmean

C均值聚类算法 ,采用修改后的聚类方法,快速的修改聚类中心,减少计算量
C mean (2010-08-15, PDF, 108KB, 下载13次)

http://www.pudn.com/Download/item/id/1269374.html

[人工智能/神经网络/深度学习] 12

启发式遗传算法在电子商务下物流配送中心选址中的研究. 电子商务环境下的物流配送中心选址问题比较复杂,涉及因素较多。根据物流配送特点,对配送中心选址模型的变动 费用和时间约束作修改,建立更加合理的适用于电子商务环境下的物流配送中心选址模型。由于选址模型属于NP难问题,故采用启 发式遗传算法进行求解。 [关键词] 电子商务 启发式遗传算法 物流配送中心
Heuristic Genetic Algorithm in e-commerce logistics distribution center location under research. E-commerce environment of logistics distribution center location problem is rather complicated, involving many factors. According to the characteristics of logistics and distribution, distribution center location model for changes in cost and time constraints to make changes to the establishment of a more reasonable environment for e-commerce logistics distribution center location model. Because of the location model are NP hard problem, so the use of heuristic genetic algorithm. [Key words] Heuristic Genetic Algorithm e-commerce logistics distribution center (2008-05-30, PDF, 364KB, 下载71次)

http://www.pudn.com/Download/item/id/477485.html

[人工智能/神经网络/深度学习] CILecture5

北京大学信息科学技术学院信息中心开设的计算智能课程,人工智能讲义_2
Beijing University College of Information Science and Technology Information Center has offered courses computational intelligence, artificial intelligence overhead _2 (2007-03-31, PDF, 749KB, 下载12次)

http://www.pudn.com/Download/item/id/261816.html

[人工智能/神经网络/深度学习] CILecture4

北京大学信息科学技术学院信息中心开设的计算智能课程,人工智能讲义_4
Beijing University College of Information Science and Technology Information Center has offered courses computational intelligence, artificial intelligence overhead _4 (2007-03-31, PDF, 528KB, 下载19次)

http://www.pudn.com/Download/item/id/261815.html

[人工智能/神经网络/深度学习] CILecture3

北京大学信息科学技术学院信息中心开设的计算智能课程,人工智能讲义_3
Beijing University College of Information Science and Technology Information Center has offered courses computational intelligence, artificial intelligence overhead Sheet_3 Can (2007-03-31, PDF, 395KB, 下载27次)

http://www.pudn.com/Download/item/id/261809.html

[人工智能/神经网络/深度学习] CIlecture2

北京大学信息科学技术学院信息中心开设的计算智能课程,人工智能讲义_2
Beijing University College of Information Science and Technology Information Center has offered courses computational intelligence, artificial intelligence overhead _2 (2007-03-31, PDF, 854KB, 下载17次)

http://www.pudn.com/Download/item/id/261808.html

[人工智能/神经网络/深度学习] CIlecture1

北京大学信息科学技术学院信息中心开设的计算智能课程,主要是人工神经网络的讲义_1
Beijing University College of Information Science and Technology Information Center has offered courses Computational Intelligence, mainly artificial neural network overhead compression (2007-03-31, PDF, 1062KB, 下载21次)

http://www.pudn.com/Download/item/id/261801.html
总计:12