一种基于类不平衡学习的情感分析方法

针对网络评论中普遍存在的负面评论较少而影响力却较大的类不平衡问题, 提出一种基于类不平衡学习的情感分析方法. 该方法利用深度学习训练过程中的概率输出, 以计算样例的信息熵作为影响因子构建交叉信息熵损失函数. 在IMDB公开数据集上进行实验验证的结果表明, 基于集成信息熵损失函数的双向长短期记忆网络能处理类不平衡问题; 对数据的统计分析结果表明, 该策略能提升基于双向长短期记忆网络的评论情感极性分类性能. 针对AUC(area under curve)指标, 使用集成信息熵损失函数的双向长短期记忆网络模型比未考虑类不平衡的深度学习模型在中位数上最多提升15.3%.     

巨灾风险大数据处理应急分类、分解、分拣算法与应用

本文主要研究巨灾风险大数据处理的应急分类、分解、分拣算法,给出了相应的算法原理和可操作的步骤.首先根据巨灾风险大数据灾害规模巨大的特征,提出了一种用来解决巨灾风险大数据中一级事件的应急分类与二级事件及以下更低级事件的应急分解算法,并以特大地震灾害作为实例进行了算法应用.接着定义了事故灾难度,用来对巨灾风险大数据处理过程中,对各种级别的事故灾难后果进行不同的数字标识.然后提出一种用来解决巨灾风险中大数据快速处理的应急分拣算法,并在汶川地震中大规模灾害的应急救援计划中进行应用.经过采用这样的应急分拣原理,就可以在面对巨灾风险大数据的复杂、繁多和零乱的重灾事件状态下,使整个应急救援方案优化,并能够有条不紊地进行救援.     

基于三维点云的岩体结构面自动分类与参数计算

结构面间距是岩体稳定性和力学特性分析中的一个重要参数，在岩石力学、采矿工程、边坡监测等领域的数值计算中广泛应用.本文以岩体边坡露头为研究对象，基于非接触测量获得的三维点云数据，提出一种基于密度聚类的结构面细化分类方法；在结构面粗略分组提取的基础上，通过投影变换、散乱点拟合等算法，求得结构面间距和岩体体积节理数.设计开发了结构面细化分类及间距等参数计算与分析原型系统，实际案例分析表明，本文方法可有效实现结构面的自动细化分类，并能够计算出间距等相关参数，可为岩体质量分级和岩体稳定性分析等提供方法支撑.     

Recent advances and challenges in materials for 3D bioprinting

3D bioprinting has been developed as a promising technology in a wide variety of biomedical applications. However, the successful implementation of bioprinting is still heavily restricted by the relatively narrow range of printable and biocompatible materials. Recently, many comprehensive reviews of the main bioprinting methods, commonly used bioprinting materials and the trending bioprinting applications have been carried out. In this review, we focused on the trends in the development of materials for 3D bioprinting. The significance and recent advances of ECM-based materials, multi-material and stimuli-responsive materials for bioprinting were summarized successively. Moreover, the challenges in current studies and future perspectives of materials for 3D bioprinting were discussed.     

基于交通信号灯信息的混合动力汽车节能预测控制方法

针对传统混合动力汽车控制方法不考虑已知道路交通信号灯信息对车辆能量管理影响的问题,提出了基于交通信号灯信息的混合动力汽车节能预测控制智能优化策略。通过建立混合动力汽车系统的简化模型,并采用连续广义最小残量方法求解模型预测控制问题。运用MATLAB/Simulink进行仿真,仿真结果验证了交通信号灯信息模型的有效性,以及所设计的模型预测控制算法大幅度提高混合动力汽车的燃油经济性的能力和实时控制性能。研究结果表明所提出的控制策略可以实现车辆行驶轨迹的优化控制,显著提高了车辆的燃油经济性,并满足系统的实时最优控制要求。     

工艺规划系统中动态模式分类的研究

提出了工艺规划系统中动态模式分类（pattern classification）的概念，介绍了模式分类判定树和动态模式分类的基本原理，并对动态模式分类的聚类质量标准进行了研究。动态模式分类是基于实例归纳（case based reasoning,CBR)的自动分类过程，具有自学习和自完善的功能。     

Solving Non-Uniqueness in Agglomerative Hierarchical Clustering Using Multidendrograms

In agglomerative hierarchical clustering, pair-group methods suffer from a problem of non-uniqueness when two or more distances between different clusters coincide during the amalgamation process. The traditional approach for solving this drawback has been to take any arbitrary criterion in order to break ties between distances, which results in different hierarchical classifications depending on the criterion followed. In this article we propose a variable-group algorithm that consists in grouping more than two clusters at the same time when ties occur. We give a tree representation for the results of the algorithm, which we call a multidendrogram, as well as a generalization of the Lance andWilliams’ formula which enables the implementation of the algorithm in a recursive way. The authors thank A. Arenas for discussion and helpful comments. This work was partially supported by DGES of the Spanish Government Project No. FIS2006–13321–C02–02 and by a grant of Universitat Rovira i Virgili.     

一类分布参数混合系统的稳定性分析

将混合系统推广到分布参数环境下,提出分布参数混合系统的概念.基于混合系统和分布参数系统的建模方法,建立了一类分布参数混合系统的模型,并对其稳定性进行分析.运用算子半群理论以及混合系统的五元组描述法及其稳定性关系比较理论的结论,通过选择适当的映射,将分布参数混合系统的的稳定性问题转化为集中参数下混合系统稳定性问题,进而得出了一致渐近稳定的充分条件并加以证明.仿真实例验证了文中结论的正确性.     

Research on consistency measurement and weight estimation approach of hybrid uncertain comparison matrix

The consistency measurement and weight estimation approach of the hybrid uncertain comparison matrix in the analytic hierarchy process （AHP） are studied. First, the decision-making satisfaction membership function is defined based on the decision making＇s allowable error. Then, the weight model based on the maximal satisfactory consistency idea is suggested, and the consistency index is put forward. Moreover, the weight distributing value model is developed to solve the decision making misleading problem since the multioptimization solutions in the former model. Finally, the weights are ranked based on the possibility degree approach to obtain the ultimate order.