算法

The paper proposes an algorithm based on the combination of Ant Colony Optimization and Particle Swarm Optimization for path planning.The new algorithm combines the advantages of ACO and PSO effectively and generates the distribution of the initial information for ACO by using the merits of high efficiency and concision of PSO,and then uses the advantages of parallelizability,positive feedback and solution with high accuracy of ACO to get global optimum solution.

提出了一种基于蚁群粒子群算法融合的机器人全局路径规划算法，该方法有效地结合了蚁群算法和粒子群算法的优点，利用粒子群算法的快速简洁等特点得到蚁群算法初始信息素分布；然后利用蚁群算法的并行性、正反馈性、求解精度高等优点，求得全局最优解。

First,we proposes the acceleration of Monte Carlo EM Algorithm,which is based on Monte Carlo EM Algorithm and Newton-Raphson algorithm,to improve the convergence rate;Second,the it is shown that the accelerated EM algorithm we proposed has quadratic convergence rate in a neighborhood of the posterior mode;Finally,its excellent performance in convergence rate is illustrated by a classical example.

受Monte Carlo EM算法与EM加速算法启发，本文构造了一种新的EM算法，称为Monte Carlo EM加速算法；证明了该算法在似然函数/后验分布的众数的附近确实具有二次收敛速度，改进了Monte Carlo EM算法的收敛速度；并通过一个数值例子的计算结果说明了该算法的优良性，它兼具实现简单及收敛速度快的特点。

First, the traffic flow time series chaotic feature is extracted by chaos theory. pretreatment for traffic flow time series, and the wavelet neural networks model was build by this. Second, the chaotic mechanism and the chaotic probability is described. Based on chaotic learning algorithm, and the wavelet neural networks fast learning algorithm of traffic flow time series is designed based on chaotic algorithm. Last, a single-step and multi-step prediction of traffic flow chaotic time series is researched by BP neural networks, wavelet neural networks and wavelet neural networks based on chaotic algorithm. The results showed that the wavelet neural networks predictive performance is better than the BP networks and the wavelet neural networks by the simulation results and root-mean-square value.

首先，通过混沌理论提取了交通流量时间序列的混沌特征，并在此基础上建立了小波神经网络交通流量时间序列模型；接着，阐述了混沌学习算法的混沌机理、混沌产生的概率，设计了基于混沌算法的小波神经网络交通流量混沌时间序列快速学习算法；最后利用交通流量混沌时间序列对BP网络、非混沌算法的小波神经网络以及基于混沌算法的小波神经网络进行了单步预测和多步预测，并对预测结果的仿真图和真实值与预测值的方均根进行了比较，结果表明基于混沌学习算法的小波神经网络的预测性能明显优于应用BP网络和非混沌算法的小波神经网络。

Then based on the analysis of running time of binary algorithm and its imroved variant, I present a parallel algorithm of scalar multiplication, and analyze the running time of that parallel algorithm.

通过分析二进制算法及作为其改进算法的2~T -ary算法、加减链算法的运行时间的影响因素，论文继而提出了将2~T -ary算法并行化的算法，并对该并行算法的运行时间进行了理论上的分析。

The basic principle and algorithm of Q-learning and related developed algorithms including Q and SARSA is studied. We distinguished between two types of RL algorithms: on-policy and off-policy, and the convergence result of SARSA(0) algorithm was discussed.

研究了Q-学习的基本原理、算法和相关的几种改进算法，如Q算法和SARSA算法；区分了两类强化学习算法：在策略和离策略算法；讨论了SARSA(0)算法的收敛性。

Considering the difference of the use of clustering algorithms caused by the diffference of data distribution, on the base of the analysis of Kmeans algorithm, fuzzy Cmeans algorithm and genetic algorithm, a new algorithm was brought forward based on the genetic algorithm and the improved neighbor function criterion. In view of the intrinsic connection of the special and the advantage of genetic algorithms wholesearch strategy, to a certian extent, this new algorithm solve the problem about the uncompact and irregular data distribution.

考虑到空间数据分布特性差异造成聚类算法采用的不同，在比较分析K均值算法、模糊均值算法和遗传算法的基础上，提出了改进的近邻函数准则，并有机整合形成遗传算法与改进近邻函数准则的新算法，综合了遗传算法的全局性概率搜索的优点，并考虑到空间数据内在的连接方式，在一定程度上较好地解决了数据的非致密非规则分布问题。

The multi-stage Tabu Search fits for the situation where the problem has middle size, the quality of solution is important, and the speed is not restricted too much.(4) The combined algorithm which shares the advantages of HR and TS—high speed and high quality—is the most practicable one for middle-scale problems.

通过仿真实验对分枝定界算法、基于运行的启发式算法、多阶段禁忌搜索算法和组合算法进行一系列的对比研究，得出以下几点结论：(1)分枝定界算法只适用于问题规模小且对计算时间要求不高的场合，它是一种最优化算法；(2)启发式算法适合于求解大规模问题，尤其是实际应用中遇到的大型问题，对计算时间要求较高而对解的质量只要满意即可。

Analyses and expatiate on the theory of MWM algorithm, which is Snorts default multi-pattern matching algorithm.

论文对几个经典的模式匹配算法的原理进行了分析研究，包括单模式匹配算法BM算法，多模式匹配算法AC算法和WM算法以及Snort系统默认使用的MWM算法。

The most essential part of the dissertation is the study on classic ID3 algorithm and Apriori algorithm. The main works are listed as follows:①The detailed theoretical background of ID3 algorithm is introduced, and based on the treeing rules. The decision tree is obtained through the practical learning of this algorithm.

本文重点对决策树方法的经典算法ID3算法和关联规则经典算法Apriori算法进行了研究，完成的主要工作如下：①在决策树方面详细介绍了经典算法ID3算法的理论背景，并按照建树的规则，通过实例使用ID3算法学习得到一棵决策树。

the effect of inertia weight on particle swarm optimization is studied,on basis of which adopts four kinds of strategies of inertia weight to regulate the speed of a new quantum delta potential well based particle swarm optimization.a faster and more stabile algorithm,found by comparing the performances of four equations regulated the inertia weight,solves 0/1 knapsack problem.the result of experiment shows that the modified algorithm improves the precision of optimal solution and has a faster speed and a higher efficiency in convergence.in a word,choosing a parameter of inertia weight suitably can improve the performance of new qdpso.

摘 要：在研究惯性权重对基本pso算法影响的基础上，根据惯性权重对粒子群算法影响的特点，采用4种惯性权重策略对一种新的具有量子行为的粒子群算法的速度进行调节，比较每种算法的性能，从中找到一种新的性能更好的改进算法，将其用于求解0-1背包问题。实验结果表明较好地选择惯性权重参数对算法的性能有很大提高，该改进算法在求解0-1背包问题中具有高效性，提高了最优解的精度，同时具有较快的收敛速度。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。