Underflow and the Denormalized Numbers

Although there have been misconceptions about it, gradual underflow fits naturally into the proposed standard and leads to simple, general statements about the arithmetic.     

基于客户端下溢概率估计的可伸缩视频传输算法

针对无线网络中可伸缩视频传输存在的问题,提出一种基于客户端下溢概率估计的传输算法.算法通过统计当前播放缓冲区容量的变化情况,计算出缓冲区下溢的概率,以此为根据选择合适的视频发送层数,从而在充分利用有效带宽的同时尽可能减少视频下溢频率.仿真结果表明,该算法能够保证在无线网络中视频传输具有较低的下溢概率和较高的视频质量.     

基于强化学习的浓密机底流浓度在线控制算法

复杂过程工业控制一直是控制应用领域研究的前沿问题. 浓密机作为一种复杂大型工业设备广泛用于冶金、采矿等领域. 由于其在运行过程中具有多变量、非线性、高时滞等特点, 浓密机的底流浓度控制技术一直是学界、工业界的研究难点与热点. 本文提出了一种基于强化学习技术的浓密机在线控制算法. 该算法在传统启发式动态规划 (Heuristic dynamic programming, HDP)算法的基础上, 设计融合了评价网络与模型网络的双网结构, 并提出了基于短期经验回放的方法用于增强评价网络的训练准确性, 实现了对浓密机底流浓度的稳定控制, 并保持控制输入稳定在设定范围之内. 最后, 通过浓密机仿真实验的方式验证了算法的有效性, 实验结果表明本文提出的方法在时间消耗、控制精度上优于其他算法.     

基于HMM的说话人识别中下溢问题的修正

基于隐马尔可夫模型构造说话人识别系统,涉及到评估、识别、训练三个基本问题,相应地采用前-后向算法、Viterbi算法、Baum-Welch算法来进行编程实现。在实现过程中,发现和的下溢问题,提出了采取增加比例因子对其加以修正的方法。     

New AspectJ Pointcuts for Integer Overflow and Underflow Detection

ABSTRACT

Aspect-oriented Programming (AOP) appears to be a promising paradigm for software security hardening. Using AOP, security experts can be responsible for coding security properties, and developers can concentrate on the basic functionality of the program. AspectJ extends the Java programming language to implement crosscutting concerns modularly in general. In this paper, we have extended AspectJ with new pointcuts in order to detect integer overflows and underflows in Java. Integer overflows and underflows in Java occur silently without throwing an exception. A malicious user can exploit them to produce a security breach. Hence, we implement new pointcuts: addition, multiplication, and subtraction that allow to write advices around integer arithmetic operations to detect integer overflow and underflow and consequently prevent considerable number of security breaches.     

基于多步下溢概率估计的自适应媒体播放算法

自适应播放(AMP)是一种通过动态调整播放速率减少这种中断的技术。多数AMP算法都基于缓存的满溢度或者其变化调整播放速率,其难点在于如何选择合适的调整门限值。本文算法,利用滑动窗口估计到达帧率的经验分布,再利用统计计算方法估计多步缓存下溢概率,然后根据一步和多步下溢概率调整播放速率。仿真实验证明在一般的信道假设和多种片源的测试中,本算法较其他同类算法有更好的表现。     

Solving Linear Interval Systems Is NP-Hard Even If We Exclude Overflow and Underflow

Reliable Computing -     

混合选别浓密机底流矿浆浓度和流量区间智能切换mm控制方法

赤铁矿磁选-浓密-浮选过程中浓密机的底流矿浆浓度受到大而频繁的浮选过程产生的中矿矿浆随机干扰的影响,造成底流矿浆流量频繁波动在工艺规定的范围之外,使得矿浆在浮选机中选别时间缩短,液位波动造成有用金属流失,从而减少精矿品位和金属回收率. 本文分析了难以采用现有的底流矿浆浓度闭环控制策略的原因,提出了由流量设定和跟踪流量设定值控制组成的矿浆浓度与流量区间双闭环控制;提出了基于静态模型的流量预设定、模糊推理的流量设定补偿、流量设定保持器和规则推理的切换机制组成的流量设定智能切换控制方法. 与矿浆浓度闭环控制方法的仿真对比实验和在国内某大型赤铁矿混合选别浓密机的成功应用,表明所提出的方法在浮选中矿干扰下,不仅将底流矿浆浓度和流量控制在目标值范围内,而且明显减少底流矿浆流量波动,从而在保证金属回收率不变的条件下,显著提高了精矿品位.     

Match-bounds revisited

The use of automata techniques to prove the termination of string rewrite systems and left-linear term rewrite systems is advocated by Geser et al. in a recent sequence of papers. We extend their work to non-left-linear rewrite systems. The key to this extension is the introduction of so-called raise rules and the use of tree automata that are not quite deterministic. Furthermore, to increase the applicability of the method we show how it can be incorporated into the dependency pair framework. To achieve this we introduce two new enrichments which take the special properties of dependency pair problems into account.     

Quicksort revisited

H.D. Mills and R.C. Linger (1986) propose adding the datatype set to existing programming languages. During some investigations using sets, it became apparent that Quicksort can be written without using stacks (or recursion). Using sets can lead to efficient multiprocessor usage, because if the elements of a set can be processed in any order, they can frequently be processed simultaneously. An example of the possibilities is an intelligent disk control unit based on External Quicksort, using four processors and four read/write heads., The control unit can sort a large disk file in about 1/3 of the time taken by the one-processor version     

Textures revisited

We describe texture generation methods for complex objects. Recently developed 3D scanning devices and high-resolution cameras can capture the complex geometry of an object and yield high-resolution images. However, generating a textured model from this input data is still a difficult problem. This task is divided into three subproblems: parameterization, texture combination, and texture restoration. A low-distortion parameterization method is presented, which minimizes geometry stretch energy. Photographs of the object taken from multiple viewpoints under modestly uncontrolled illumination conditions are merged into a seamless texture using our new texture combination method. We also demonstrate a texture restoration method that can fill in missing pixel information when the input photographs do not provide sufficient information to cover the entire surface due to self-occlusion or registration errors. Our methods are fully automatic, except for the registration process between a 3D model and input photographs. We demonstrate the application of our method to human face models for evaluation. The techniques presented in this paper make a consistent and complete pipeline to generate the texture of a complex object.     

Redundancy reduction revisited

Soon after Shannon defined the concept of redundancy it was suggested that it gave insight into mechanisms of sensory processing, perception, intelligence and inference. Can we now judge whether there is anything in this idea, and can we see where it should direct our thinking? This paper argues that the original hypothesis was wrong in over-emphasizing the role of compressive coding and economy in neuron numbers, but right in drawing attention to the importance of redundancy. Furthermore there is a clear direction in which it now points, namely to the overwhelming importance of probabilities and statistics in neuroscience. The brain has to decide upon actions in a competitive, chance-driven world, and to do this well it must know about and exploit the non-random probabilities and interdependences of objects and events signalled by sensory messages. These are particularly relevant for Bayesian calculations of the optimum course of action. Instead of thinking of neural representations as transformations of stimulus energies, we should regard them as approximate estimates of the probable truths of hypotheses about the current environment, for these are the quantities required by a probabilistic brain working on Bayesian principles.     

The backchase revisited

Semantic query optimization is the process of finding equivalent rewritings of an input query given constraints that hold in a database instance. In this paper, we report about a Chase & Backchase (C&B) algorithm strategy that generalizes and improves on well-known methods in the field. The implementation of our approach, the Pegasussystem, outperforms existing C&B systems an average by two orders of magnitude. This gain in performance is due to a combination of novel methods that lower the complexity in practical situations significantly.     

Bin stretching revisited

We study three on-line models of bin stretching on two machines. For the case where the machines are identical and the jobs arrive sorted by non-increasing sizes, we show a tight bound of 10/9 on the competitive ratio. For two related machines, we show a preemptive algorithm with competitive ratio 1 for any speed ratio, and two new non-preemptive algorithms. We prove that the upper bound on the competitive ratio achieved by the non-preemptive algorithms is optimal for almost any speed ratio, and close to optimal for all other speed ratios. Received: 14 February 2002 / 18 November 2002 Research supported in part by the Israel Science Foundation, (grant No. 250/01-1).     

Surface skinning revisited

Published online: 15 March 2002