拟概率空间上结构风险最小化原则

首先,给出了拟概率空间上结构风险最小化原则。然后,为了解决在拟概率空间上结构风险是否一致收敛到期望风险,也就是根据这个最小化原则结构风险是否能收敛到最小可能的风险,给出并证明了结构风险最小化原则的一致收敛性。     

一种SVM验证码识别算法

设计验证码的主要目的是区分人类和计算机,用来防止网络机器人的一些恶意行为。验证码的出现也催生了一批新工种,电子商务的发展迫切需要一种推广方式来推销他们的商品,残障人士上网问题也需要迫切关注,因此许多人开始研究网络机器人技术,用来实现邮箱自动注册、群发信息、自动灌水、自动登录等功能。目前,各种类型网站系统都利用验证码阻止网络机器人入侵,从而验证码识别技术成为研究热点。基于SVM技术对图像验证码进行识别,取得了良好的效果。     

三门核电站放射性固体废物管理

介绍了三门核电站放射性固体废物的处理工艺,以及放射性固体废物管理的思路,在参考国内其他核电站放射性固体废物管理经验的基础上,对三门核电站放射性固体废物最小化方法进行初步探索。     

A Bayes Sampling Allocation Scheme for Stratified Finite Populations With Hyperbinomial Prior Distributions

When sampling is carried out independently for the K strata of a finite stratified dichotomous population (defectives vs. standard items), and the number Z_i of defectives per stratum sample is observed, the corresponding probability function for X = (X_i , …, x_K ) is the product of hypergeometric functions which depend on the sample sizes n_i , the stratum sizes N_i , and the number of defectives M_i in the stratum (i = 1, …, K). It is assumed that prior information is available about the M_i 's which can be expressed, by suitable choice of the parameters a_i and b_i , as the product of independent hyperbinomial functions.

In each stratum the cost per observation is a known constant. Using squared error loss function, the prior Bayes risk is found for the linear function of interest,

and the optimum allocation of sample sizes is found, the one for which the prior Bayes risk is minimum when the total sampling budget is fixed.     

放射性废物最小化管理系统初步设计

利用面向对象的软件工程方法,针对中国原子能科学研究院的废物管理,开发设计了放射性废物最小化管理系统.本文介绍该系统的开发背景、近期目标、系统的功能、系统的数据流分析、数据表设计与功能说明,以及系统的主要算法等.     

苏丹膨胀土的工程性质与处理方法

以苏丹Munga地区某输电线路岩土工程勘察项目为背景,分析了该区膨胀土的工程性质,并针对该区膨胀土的胀缩性能提出了膨胀土危害的防灾思路,建立了该区膨胀土塑性指数与液限的相关方程,对膨胀力与初始含水量的相关性进行了分析,研究成果对该地区工程建设地基基础设计及我国膨胀土的研究具有一定参考价值。     

Equilibrium modeling of gasification: Gibbs free energy minimization approach and its application to spouted bed and spout-fluid bed gasifiers

Spouted beds have been found in many applications, one of which is gasification. In this paper, the gasification processes of conventional and modified spouted bed gasifiers were considered. The conventional spouted bed is a central jet spouted bed, while the modified spouted beds are circular split spouted bed and spout-fluid bed. The Gibbs free energy minimization method was used to predict the composition of the producer gas. The major six components, CO, CO₂, CH₄, H₂O, H₂ and N₂, were determined in the mixture of the producer gas. The results showed that the carbon conversion in the gasification process plays an important role in the model. A modified model was developed by considering the carbon conversion in the constraint equations and in the energy balance calculation. The results from the modified model showed improvements. The higher heating values (HHV) were also calculated and compared with the ones from experiments. The agreements of the calculated and experimental values of HHV, especially in the case of the circular split spouted bed and the spout-fluid bed were observed.     

Computation of logical effort in high level languages

A new measure of software complexity is introduced, that of logical effort. This measure is an attempt to quantify program complexity by an analysis of the control structure of the program, using the concepts of language virtuality and segment independence. A program is decomposed by the use of these concepts until measurements can be made at the statement level. These measurements, along with measurements of I, the effort necessary to perform a loop, and Q, the effort necessary to determine which branch to take in a decision, are then used to calculate recursively the effort in larger and larger structures, until the program level is reached. An example of this process is given for a short program, along with a brief comparison of the results obtained with a similar measure, that of cyclomatic complexity.

Algorithms for the computation of I and Q are also given, along with examples of their calculation. This process is essentially a decomposition of a Boolean function into disjunctive normal form, followed by a minimization of the form using a weighting measure involving weights on both the operators and the variables in the function.     

基于多优化目标的软件测试用例约简方法研究

软件测试是软件工程的一个重要组成部分,其目标是能够及时发现软件中的错误,确保软件高质量。测试用例是软件测试的基础,覆盖度较高且精简的测试用例集可以提高测试效率和降低成本。软件测试覆盖标准较多,一个好的测试用例评价指标也存在多种,为了能够在约简测试用例集规模的同时获取较高的测试能力,本文提出了一种基于多优化目标的测试用例集约简算法,该算法旨在根据测试用例需求,构建多优化目标的测试用例模型,使用该模型获取一个最优解的测试用例子集,使用最小化用例集方法最小化测试用例,迭代执行直到测试用例集覆盖所有的测试需求,实验结果表明该算法可以约简测试用例集,获取较高的综合测试效果。