多类分类SVM在工程车辆自动变速挡位决策中的应用

经典的支持向量机(SVM)是针对二类分类的,在解决工程车辆自动变速挡位决策这种典型的多类分类问题时存在困难。本文提出了基于二叉数支持向量机的挡位决策算法,将分类器分布在各个节点上,从而构成了多类分类支持向量机,减少了分类器数量和重复训练样本的数量。该方法能够根据车辆的运行状态确定最佳挡位,从而及时、准确地满足工程车辆自动换挡的要求。试验结果表明基于二叉树的支持向量机性能要比遗传RBF神经网络略好。

集成环境下基于实例的零件工艺设计

在零件特征信息描述基础上,采用二叉数描述零件信息的数据结构,针对基于实例的回转体零件工艺设计,提出了基于二叉数遍历及搜索的实例检索算法以及实例修改算法。由于采用了统一的数据模型,实现了CAD/CAPP系统的信息集成。

Paired bialgebras and braided bialgebras

First, we present semisimple properties of twisted products by means of constructing an algebra isomorphism between twisted products and crossed products, and point out that there exist some relations among braided bialgebras, paired bialgebras and Yang-Baxter coalgebras. Furthermore, we give an example to illustrate these relations by using Sweedler's 4-dimensional Hopf algebra. Finally, from starting off with Yang-Baxter coalgebras, we can construct some quadratic bialgebras such that they are braided bialgebras.

Research on Anti-collision Algorithm of Short Distance Data communication based on Binary-Tree Disassembly

This paper puts forward adaptive anti collision algorithm based on two fork tree decomposition. New search algorithm built on the basis of binary-tree algorithm, using the uniqueness of the label EPC, to estimate the distribution of label by slot allocation, the huge and complicated two fork tree is decomposed into several simple binary-tree by search the collision slots for binary-tree, so, it can simplifies the search process. The algorithm fully considers4 important performance parameters of the reader paging times, transmission delay, energy consumption and throughput label, the simulation results show that, the improved anti-collision algorithm is obviously improved performance than other two fork tree algorithm, it is more suitable for RFID anti-collision protocols.

浅谈数据结构——树

本文简单概述了树,二叉树以及哈夫曼树的基本概念,针对树的基本应用做了简要的说明,以及树的这种存储数据结构在数据存储中的时间复杂度和空间复杂度的简要分析。