PRISM – A novel framework for pattern recognition

In this paper, we introduce a new model of solving pattern recognition tasks called PRISM (Pattern Recognition using Information Slicing Method). The main concept behind PRISM is the slicing of information through multiple planes across different feature axes to generate a number of cells. The number of cells created and their volume depends upon the number partitions per axes. In this context we define resolution as the number of partitions per axes. In this paper, we make the following contributions. First, we provide a brief survey of the class separability measures and feature partitioning schemes used for pattern recognition. Secondly, we define the PRISM framework and the algorithm for data assignment to cells. Thirdly, we detail four important concepts in PRISM: purity, neighbourhood separability, collective entropy, and data compactness. The first two measures define the data complexity, the next measure relates to uncertainty, and the last measure defines the alternative to statistical data variance in the PRISM framework. Fourthly, we investigate the variability in the estimates of these measures depending on the placement of partitions on each feature axis. Finally, we give an overview of experimental successes achieved with PRISM in the areas of classification complexity estimation and feature selection.     

一种高效的动态概率广播算法及其概率模型检测分析

在Ad Hoc网络中,广播有着相当广泛的应用,其算法的效率极大地影响着网络的性能.本文基于DP算法提出了BN-DP算法,考虑了节点分布、计数器值以及收发节点间距离对广播算法的影响,赋予处于接收边缘的节点更高的转发概率.然后使用概率模型检测工具PRISM,分析了计数器值和节点分布对BN-DP算法性能的影响.结果表明:在相同可达率的情况下,所提出的BN-DP算法与FP、DP算法相比,减少了转发分组的数量,提高了广播效率.     

用PTA模型形式化分析基于Gossip 协议的发布/订阅系统

在研究传统的发布／订阅消息中间件系统的基础之上,结合Gossip协议的特点来研究发布／订阅消息中间件,最后运用形式化方法,通过PRISM仿真工具,对仿真的模型进行形式化分析．实验结果表明,发布／订阅消息中间件系统的实时性受消息产生速度的影响,在各个订阅者订阅相同消息和不同消息两种情况之下网络特性展现不同的变化,但最终都是随着消息产生速度的增加而减小．可靠性随着消息产生速度的增加而减小,并且订阅者的接收缓存越大可靠性越高,但增幅率会越来越小．该实验模型和实验方法对于发布／订阅消息中间件系统的研究,以及在现实环境中配置系统的相关参数有一定的帮助．     

一个分布式K互斥算法的概率模型检测

传统的验证方法难以保证分布式K互斥算法的有效性和安全性.为解决这一问题,给出了进一步的研究,提出一种基于概率模型检测器PRISM的方法,对Kerry Raymond的分布式K互斥算法进行形式化建模与分析验证.通过设置算法中各个进程进入临界区的时间而得出的结果中发现,改变临界区的数目K,对于某一进程进入临界区的平均及时时间的影响并不大.如果某一进程的执行时间比其他进程大很多,则K的增加可以提高运行效率.最后证明了这一结论.     

基于概率模型检测的合同签订协议分析

概率合同签订协议的公平性具有典型的概率性质。概率模型检测是一种验证存在随机行为系统的分析技术。基于概率模型检测技术，在原协议的基础上建立了离散马尔可夫链和有限状态机模型，用概率模型检测工具——PRISM验证了它的公平性，发现其公平性不满足时，在此基础上对原协议从双方交换信息的策略上进行了扩充改进，使协议更为公平，满足了合同签订的基本要求。     

基于模型检测的微服务组合平台QoS验证

针对当前缺少对微服务组合平台的服务质量（QoS）指标进行分析验证的问题，提出一种基于模型检测的形式化验证方法，对影响微服务平台性能的因素进行分析评估。首先，将微服务组合的服务资源配置过程划分为服务请求、资源配置和服务执行3个阶段，并分别由服务请求队列、服务请求配置器和提供服务资源的虚拟机等模块实现；其次，将各个模块的实现过程建模为带标记Markov回报模型（LMRM），借助类似于进程代数的同步概念获得微服务组合过程的全局模型；接着，用连续随机回报逻辑公式刻画期望的QoS指标；最后，将形式模型与逻辑公式作为模型检测工具PRISM的输入以获得验证结果。实验结果表明，LMRM可较好地用于微服务组合平台的建模和QoS验证分析。     

从单视点DVC到多视点DVC的研究进展

多视点视频编码对编码器的功耗、复杂度、存储能力和传输能力有较高的要求,而分布式信源编码理论恰恰适应了这一需求.为了使人们对这一技术现状有一概略了解,首先总结了分布式视频编码的基本理论和关键技术,然后介绍了分布式信源编码理论从单视点到多视点的拓展,接着论述了分布式多视点视频编码的研究现状,最后分析了分布式信源编码理论在多视点视频编码领域的发展前景.     

基于概率模型检测的合同签订协议分析

概率合同签订协议的公平性具有典型的概率性质。概率模型检测是一种验证存在随机行为系 统的分析技术。基于概率模型检测技术,在原协议的基础上建立了离散马尔可夫链和有限状态机模型, 用概率模型检测工具——PRISM验证了它的公平性,发现其公平性不满足时,在此基础上对原协议从 双方交换信息的策略上进行了扩充改进,使协议更为公平,满足了合同签订的基本要求。     

Reduction of a detailed kinetic model for the ignition of methane/propane mixtures at gas turbine conditions using simulation error minimization methods

Natural gas is the primary fuel for industrial gas turbines, which provide about one quarter of the world’s primary energy supply. Beside methane it also contains larger hydrocarbons in small, varying ratios. This variation is expected to rise due to the increasing usage of non-traditional gas sources. Fuel composition has a large impact on auto-ignition delay time, which is a fundamental parameter for the optimal design and operation of gas turbines. For the oxidation of such mixtures, Curran, Petersen and co-workers recently developed a detailed reaction mechanism (NUIG NGM), which reproduces the ignition delays over a wide range of conditions. However, due to its large size: 229 species and 1359 reactions, it cannot be used in computational fluid dynamics simulations, which is an important fundamental tool in the development of gas turbines. A mechanism reduction case study of the NUIG NGM is presented using the recently developed simulation error minimization methods (SEM). A new version of the SEM program package is also proposed, which allows the reduction of mechanisms for a wider range of combustion phenomena. Combinational strategies have been introduced in the SEM connectivity method to enhance the reduction procedure and a hierarchical reduction procedure is proposed for multi-scenario problems. Ignition of lean and stoichiometric mixtures containing 90% methane and 10% propane as fuel were investigated for 22 conditions relevant to gas turbines, covering temperature and pressure ranges of 877–1465 K and 7–40 atm, respectively. The smallest reduced mechanism developed contains 50 species and 186 reactions. It can reproduce ignition delays with 3.1% maximum error and reproduces pressure rise precisely (error∼10⁻³%). The mechanism can be simulated 62 times faster than the full mechanism. Robustness analysis showed that it is reliably applicable over a much wider range of conditions compared to that for which it was developed.     

A glimpse of symbolic-statistical modeling by PRISM

We give a brief overview of a logic-based symbolic modeling language PRISM which provides a unified approach to generative probabilistic models including Bayesian networks, hidden Markov models and probabilistic context free grammars. We include some experimental result with a probabilistic context free grammar extracted from the Penn Treebank. We also show EM learning of a probabilistic context free graph grammar as an example of exploring a new area.