基于预测对抗网络的图像二分类模型

正未标记学习仅使用无标签样本和正样本训练一个二分类器, 而生成式对抗网络(generative adversarial networks, GAN)中通过对抗性训练得到一个图像生成器. 为将GAN的对抗训练方法迁移到正未标记学习中以提升正未标记学习的效果, 可将GAN中的生成器替换为分类器C, 在无标签数据集中挑选样本以欺骗判别器D, 对C与D进行迭代优化. 本文提出基于以Jensen-Shannon散度(JS散度)为目标函数的JS-PAN模型. 最后, 结合数据分布特点及现状需求, 说明了PAN模型在医疗诊断图像二分类应用的合理性及高性能. 在MNIST, CIFAR-10数据集上的实验结果显示: KL-PAN模型与同类正未标记学习模型对比有更高的精确度(ACC)及F1-score; 对称化改进后, JS-PAN模型在两个指标上均有所提升, 因此JS-PAN模型的提出更具有合理性. 在Med-MNIST的3个子图像数据集上的实验显示: KL-PAN模型与4个benchmark有监督模型有几乎相同的ACC, JS-PAN也有更高表现. 因此, 综合PAN模型的出色分类效果及医疗诊断数据的分布特征, PAN作为半监督学习方法可获得更快、更好的效果, 在医学图像的二分类的任务上具有更高的性能.     

Multi-coefficients goal programming

In today’s increasingly competitive business environment, maintaining profit margins or quantities of goods sold is an important issue for businesses. Accordingly, more and more industries use group pricing discrimination strategy to attract potential customers in order to increase competitive advantage. Hence, to find ways maximize profit and to minimize total cost, group pricing discrimination strategy has become an important issue for decision makers. Unfortunately, these types of problems cannot be solved by any current goal programming models. The objective of this study is to deduce a new method, which we call the multi-coefficients goal programming, for group pricing discrimination problems. In addition, an example is given to illustrate the correctness and usefulness of the proposed model.     

Event vertex and time reconstruction in large-volume liquid scintillator detectors

Large-volume liquid scintillator detectors with ultra-low background levels have been widely used to study neutrino physics and search for dark matter. Event ve...     

郑州市机关信息系统

郑州市机关信息系统是一个以郑州市委市政府两个办公厅为核心的、基于ATM主干交换网络的、具有Intranet模式的现代化办公系统。该文结合郑州市机关信息系统的设计与实现,给出了适合机关网络建设和信息服务的一种设计和实现方法。     

南茶高纬度北引对茶叶品质的影响

目的研究南茶高纬度北引后对茶类适制性及绿茶品质产生的影响。方法通过理化方法,分析同一品种茶树在高纬度北引后对其叶片厚度、鲜叶成分、不同茶类感官品质、绿茶成分及感官品质所产生的变化。结果随着纬度的增加,绿茶品质较好,而红茶品质较差;随着南茶高纬度北引,叶片厚度、氨基酸、水溶性糖、水浸出物和芳香物质(精油)总量等含量随纬度增大而增加;茶多酚含量则相反;对精油中各成分分析显示,北方茶烃类成分较多,醇类成分较少。结论本研究明确了南茶高纬度北引后茶叶品质特点及化学成分基础。     

瞄准线稳定精度的室内测试技术

系统地介绍了瞄准线稳定精度的测量原理和测量系统组成,对测量系统中的主要技术如环境模拟、图像处理及数据拟合等技术进行了详细的描述。文中还阐述了测量方法、步骤及为提高测量精度采用的标定方法。最后给出了一个典型的测量实例。     

四川泡菜发酵过程中大肠杆菌消长规律的定量系统评价

目的探究四川泡菜发酵过程中大肠杆菌的消长规律。方法根据相关关键词在CNKI数据库中检索相关文献。对检索结果 13篇文献的50批次四川泡菜发酵过程中大肠杆菌消长数据进行定量系统评价。结果四川泡菜发酵过程中,大肠杆菌在数量上大体呈现为前升后降的动态变化模式;尽管乳酸菌生长及发酵产酸总体上对大肠杆菌产生抑制作用,但发酵过程中的多种参数和因素对大肠杆菌的消长具有显著影响。此外,泡菜发酵中,大肠杆菌数量的峰值与初始值之间有显著的正向相关关系;而发酵温度与泡菜发酵中大肠杆菌消失时间之间呈现反向相关关系。结论本研究可为进一步探讨四川泡菜发酵对大肠杆菌的抑制机制及风险评估提供参考。     

CAD技术在输电线路风偏校核中的应用

本文研究运用CAD图解法进行塔头间隙风偏校核的新思路,克服了传统解析法计算空间距离困难和图解法精度不高的缺点,并以500 kV凤磁Ⅱ回线路部分杆塔为实例,在计算悬垂串绝缘子摇摆角的基础上,利用三维制图软件AutoCAD对酒杯塔和ZV1型拉线塔进行风偏校核,分析了采用合成绝缘子取代普通玻璃绝缘子的可行性.     

A hybrid session key exchange algorithm for highly-sensitive IP-based institutional communications

Microsystem Technologies - Session key exchange has become an addressed topic in the field of communications security, particularly for the IP-based call session that travels through the public...     

Adaptive extended state space predictive control for a kind of nonlinear systems

This paper presents an adaptive nonlinear predictive control design strategy for a kind of nonlinear systems with output feedback coupling and results in the improvement of regulatory capacity for reference tracking, robustness and disturbance rejection. The nonlinear system is first transformed into an equal time-variant system by analyzing the nonlinear part. Then an extended state space predictive controller with a similar structure of a PI optimal regulator and with P-step setpoint feedforward control is designed. Because changes of the system state variables are considered in the objective function, the control performance is superior to conventional state space predictive control designs which only consider the predicted output errors. The proposed method is tested and compared with latest methods in literature. Tracking performance, robustness and disturbance rejection are improved.