Development of a safety critical software requirements verification method with combined CPN and PVS: a nuclear power plant protection system application

Safety-critical software systems such as certain nuclear instrumentation and control (NI&C) systems should be developed with thorough verification. This study presents a method of software requirement verification with a case study for a nuclear power plant (NPP) protection system. The verification introduces colored petri net (CPN) for system modeling and prototype verification system (PVS) for mathematical verification. In order to aid flow-through from modeling by CPN to mathematical proof by PVS, an information extractor from CPN models has been developed in this paper. In order to convert the extracted information to the PVS specification language, a translator has also been developed. This combined method has been applied to the functional requirements of the Wolsong NPP Shut Down System #2 (SDS2); logical properties of the requirements were verified. Through this research, guidelines and a tool support for the use of formal methods have been developed for application to NI&C software verification.     

High training set size reduction by space partitioning and prototype abstraction

Instance-based learning methods like the nearest neighbour classifier generally suffer from the indiscriminate storage of all training instances, resulting in large memory requirements and slow execution speed. In this paper, new training set size reduction methods based on prototype generation and space partitioning are proposed. Experimental results show that the new algorithms achieve a very high reduction rate with still an important classification accuracy.     

Remarks on different reviews of Chinese character recognition

This paper gives an introduction and remarks on two review papers for Chinese character recognition. One review is made by Chinese authors, another is from American scientists. They investigate Chinese character from different language environments; they do the research from different points of view. Thus, a more comprehensive view on Chinese character recognition, which is an important branch of pattern recognition, can be provided to the readers. Meantime, one article pays attention to online process, and other paper deals with offline recognition, which complement each other. The author is the Associate Editor-in-Chief of Frontiers of Computer Science in China     

Vote counting measures for ensemble classifiers

Various measures, such as Margin and Bias/Variance, have been proposed with the aim of gaining a better understanding of why Multiple Classifier Systems (MCS) perform as well as they do. While these measures provide different perspectives for MCS analysis, it is not clear how to use them for MCS design. In this paper a different measure based on a spectral representation is proposed for two-class problems. It incorporates terms representing positive and negative correlation of pairs of training patterns with respect to class labels. Experiments employing MLP base classifiers, in which parameters are fixed but systematically varied, demonstrate the sensitivity of the proposed measure to base classifier complexity.     

Comparison among feature extraction methods for HIV-1 protease cleavage site prediction

Recently, several works have approached the HIV-1 protease specificity problem by applying a number of methods from the field of machine learning. However, it is still difficult for researchers to choose the best method due to the lack of an effective comparison. For the first time we have made an extensive study on methods for feature extraction for the problem of HIV-1 protease. We show that a fusion of classifiers trained in different feature spaces permits to obtain a drastically error reduction with respect to the performance of the state-of-the-art.     

A clustering method for automatic biometric template selection

The problem addressed in this paper is the template selection and update in biometrics based on clustering. Template selection is a reliable method to reduce the number of templates used in a biometric system to account for variations observed in a person's biometric data. An efficient method based on clustering with automatic selection of the number of clusters is proposed in this work for finding subgroups of similar templates which are used for prototype selection.Experimental results confirm the advantage of the new method and the importance of adopting a procedure to perform template selection.     

基于特征选择的集成分类器抗噪性能分析

特征选择有助于增强集成分类器成员间的随机差异性,从而提高泛化精度。研究了随机子空间法（RandomSub-space）和旋转森林法（RotationForest）两种基于特征选择的集成分类器构造算法,分析讨论了两算法特征选择的方式与随机差异程度之间的关系。通过对UCI数据集引入噪声,比较两者在噪声环境下的分类精度。实验结果表明：当噪声增加及特征关联度下降时,基本学习算法及噪声程度对集成效果均有影响,当噪声增强到一定程度后。集成效果和单分类器的性能趋于一致。     

面向目标的带先验概率的AdaBoost算法

针对集成学习算法研究中多个分类器的最佳组合问题,改进了传统的AdaBoost集成学习算法.用于组合的各个分类器通常是基于样本集通过一定的训练得到,样本集中不同类目标的比率可以反映分类目标的先验概率.使用该参数给出了新的组合参数和投票表决阈值计算公式,巧妙的利用样本权值并将其加入到样本属性上进行训练学习,采用新的策略来选择基分类器,给出了面向目标的带先验概率的AdaBoost算法(GWPP AdaBoost算法)和分类器的最佳组合.依据UCI实验数据对传统的AdaBoost算法、Bagging算法、GWPP AdaBoost算法的错误率和性能进行了比较分析,验证了GWPP AdaBoost的有效性.