Applications of SMP Bounds to Multi-class Traffic in High-speed Networks

In this paper, we consider the stochastic fluid-flow model of a single node in a high-speed telecommunication network handling multi-class traffic. The node has multiple buffers, one for each class of traffic. The contents of these buffers are multiplexed onto a single output channel using one of the service scheduling policies: the Timed Round Robin Policy or the Static Priority Service Policy. The Quality of Service requirements for each class are based on cell loss probabilities. Using effective bandwidth methodologies and the recently developed bounds for semi-Markov modulated traffic, we solve call admission control problems for the two service scheduling policies at this node. We compare the performance of the effective bandwidth methodologies and the SMP bounds technique. We also numerically compare the performance of the two service scheduling policies.     

AIS—NC多类型开放队列网络性能分析

本文研究了基于网络计算机会计信息系统性能分析方法，针对会计信息系统结构特点及硬件资源请求类型，抽象出三种系统响应模式．据此，构建了AIS-NC多类型开放队列网络性能模型，并且给出了性能分析的步骤和方法．举例说明了该方法的应用．该方法对其他多类型开放队列网络性能分析具有借鉴意义．     

模糊多类SVM模型

利用SVM处理多类分类问题,是当前的研究热点之一.本文提出了一种模糊多类支持向量机模型,即FMSVM.该方法是在Weston等人提出的多类SVM模型中引入模糊成员函数,针对每个输入数据对分类结果的不同影响,该模糊成员函数得到相应的值,由此得到不同的惩罚值.从而在构造分类超平面时,可以忽略那些对分类结果影响很小的数据.理论分析与数值实验都表明,该算法具有良好的鲁棒性.     

ERM learning algorithm for multi-class classification

The multi-class classification problem is considered by an empirical risk minimization (ERM) approach. The hypothesis space for the learning algorithm is taken to be a ball of a Banach space of continuous functions. When the regression function lies in some interpolation space, satisfactory learning rates for the excess misclassification error are provided in terms of covering numbers of the unit ball of the Banach space. A comparison theorem is proved and is used to bound the excess misclassification error by means of the excess generalization error.     

连续型异质有限理性用户分类及其随机网络均衡

为提高交通行为引导政策的针对性,有必要研究随机路网环境下不同类型出行者对交通引导政策的反应,即多类有限理性用户风险态度和路网退化系数对用户行为决策的影响.本文构建了连续风险态度离散化的用户分类模型以及基于退化路网的多类有限理性用户均衡模型,证明了解的存在性,并设计了启发式算法求解.基于双向随机退化Nguyen&Dupuis路网模拟了10类用户同时出行的均衡状态.分析了均衡时异质用户行程时间期望、标准差和累积到达时间感知价值等参数与用户风险态度和路网退化程度之间的关系.研究发现行程时间期望、标准差与其用户类别并无显著差异;行程时间标准差与退化路网随机性正相关,但未发现行程时间期望与路网随机性相关.用户累积到达时间感知价值随风险态度增加而递减,即保守用户感受到的累积到达时间价值更小.因此在相同的引导策略下,保守用户的广义效用函数可能变化更大,政策引导效果可能更加显著.     

Workloads and waiting times in single-server systems with multiple customer classes

One of the most fundamental properties that single-server multi-class service systems may possess is the property of work conservation. Under certain restrictions, the work conservation property gives rise to a conservation law for mean waiting times, i.e., a linear relation between the mean waiting times of the various classes of customers. This paper is devoted to single-server multi-class service systems in which work conservation is violated in the sense that the server's activities may be interrupted although work is still present. For a large class of such systems with interruptions, a decomposition of the amount of work into two independent components is obtained; one of these components is the amount of work in the corresponding systemwithout interruptions. The work decomposition gives rise to a (pseudo)conservation law for mean waiting times, just as work conservation did for the system without interruptions.     

一种基于HDDT集成的多类不平衡学习方法

在很多真实世界问题中,不同类别的数据样本往往有显著的不平衡性,即大类的样本远多于小类.对类别不平衡样本进行学习,是目前国内外数据挖掘和机器学习领域的研究热点之一.以往对不平衡样本学习的研究主要针对二分类问题进行,由此针对多分类问题,提出一种基于HDDT决策树集成的多类不平衡学习方法.实验表明,该方法可以有效地对多类不平衡问题进行学习.     

支撑向量机多类分类方法的研究

支撑向量机是一种基于统计学理论的新的学习算法,它采用了结构风险最小化原则,能有效地解决过学习问题,具有很强的泛化能力。传统支撑向量机针对两类分类问题,为了深入地分析实际应用中的大规模和多类别的问题,通过对一次性求解法、一对多、一对一、有向无环图方法的原理和实现方法进行分析,从速度和精度两方面对这些方法的优缺点进行了归纳和总结,并通过实验进行了验证和比较。实验结果表明,各种方法可以获得不同的分类器推广能力及训练速度和测试速度,也为今后如何更好地解决支撑向量机多类分类问题指明了方向。     

The Dependence of Optimal Returns from Multi-class Queueing Systems on Their Customer Base

We identify structured collections of multi-class queueing systems whose optimal return (a minimised cost) is a supermodular function of the set of customer classes allowed external access to the system. Our results extend considerably the range of systems for which such a claim can be made. The returns from such systems also exhibit a form of directional convexity when viewed as functions of a vector of arrival rates. Applications to load balancing problems are indicated.     

Multi-Class Procedure for Analysis of 50 Antibacterial Compounds in Eggshells Using Ultra-High-Performance Liquid Chromatography–Tandem Mass Spectrometry

In this work, for the first time, Ultra-High-Performance Liquid Chromatography–Tandem Mass Spectrometry (UHPLC–MS/MS) method was developed for qualitative and quantitative analysis of veterinary antibiotics (cephalosporins, diaminopyrimidines, fluoro(quinolones), lincosamides, macrolides, penicillins, pleuromutilins, sulfonamides, tetracyclines, and sulfones) in hen eggshells. The sample preparation method is based on a liquid–liquid extraction with a mixture of metaphosphoric acid, ascorbic acid, EDTA disodium salt dihydrate, and acetonitrile. The chromatographic separation was performed on Luna^® Omega Polar C18 10 column in gradient elution mode and quantitated in an 8 min run. Validation such as linearity, selectivity, precision, recovery, matrix effect, limit of quantification (LOQ), and limit of detection (LOD) was found to be within the acceptance criteria of the validation guidelines of the Commission Decision 2002/657/EC and EUR 28099 EN. Average recoveries ranged from 81–120%. The calculated LOQ values ranged from 1 to 10 µg/kg, the LOD values ranged from 0.3 to 4.0 µg/kg, depending on analyte. The developed method has been successfully applied to the determination of antibacterial compounds in hen eggshell samples obtained from different sources. The results revealed that enrofloxacin, lincomycin, doxycycline, and oxytetracycline were detected in hen eggshell samples.