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.     

利用反预报方法从截面预言近滴线丰质子核的结合能

在Origin等数据处理软件中,当实验数据较少时,自由参数的不同初始化设置会导致较大的结果差异,这为物理结果的确定带来较大不确定性.通过最小二乘法分析了345MeV/u ~(78) Kr+~9Be反应中产生的丰质子同位素的截面和结合能,并得到线性回归方程.通过回归方程,利用结合能预报部分丰质子核素的截面,以及通过实验截面对近质子滴线核素的结合能进行反预报测量.这对于近质子滴线的丰质子核素实验测量具有较好的借鉴意义.     

基于变量选择-神经网络模型的复杂路网短时交通流预测

摘要： 针对传统交通流预测模型正在由单断面历史数据处理向多断面、多时刻历史数据处理转变,但在考虑各断面间的影响时,多变的交通状况往往会使预测模型复杂化的问题,引入一种多元线性回归最小绝对收缩和选择算子方法(Lasso),并利用其优秀的变量选择能力,在复杂路网多断面中选出相关性较高的断面；结合神经网络(NN)的非线性特性,提出了Lasso NN组合模型.结果表明：Lasso NN模型在路网交叉口对未来15 min交通流数据预测的误差率低于9.2%；在非交叉口的误差率低于6.7%,总体优于各自单独使用得出的结果.     

线性回归分析与能源需求预测

结合实例介绍了回归模型方法在能源预测中的应用，具体地讨论了最简单、最基本的直线回归模型结构及参数估计方法，对于其他一些曲线回归模型则可通过变量代换转化为直线回归模型。     

Robustness of alternative non‐linearity tests for SETAR models

In recent years there has been a growing interest in exploiting potential forecast gains from the non‐linear structure of self‐exciting threshold autoregressive (SETAR) models. Statistical tests have been proposed in the literature to help analysts check for the presence of SETAR‐type non‐linearities in an observed time series. It is important to study the power and robustness properties of these tests since erroneous test results might lead to misspecified prediction problems. In this paper we investigate the robustness properties of several commonly used non‐linearity tests. Both the robustness with respect to outlying observations and the robustness with respect to model specification are considered. The power comparison of these testing procedures is carried out using Monte Carlo simulation. The results indicate that all of the existing tests are not robust to outliers and model misspecification. Finally, an empirical application applies the statistical tests to stock market returns of the four little dragons (Hong Kong, South Korea, Singapore and Taiwan) in East Asia. The non‐linearity tests fail to provide consistent conclusions most of the time. The results in this article stress the need for a more robust test for SETAR‐type non‐linearity in time series analysis and forecasting. Copyright © 2004 John Wiley & Sons, Ltd.     

动态校正的PI型广义预测控制器及其在反应釜系统中的仿真研究

提出了一种新型的广义预测控制器。由于在控制器的设计中,进行了预测误差的动态校正,并且控制器采用了比例积分型(PI)目标函数,因此该控制器对未建模动态具有很强的鲁棒性。应用低阶的模型对一个强放热的化学反应釜模型进行的仿真研究检验了该方法的有效性。     

Finding finer functions for partially characterized proteins by protein-protein interaction networks

Based on high-throughput data, numerous algorithms have been designed to find functions of novel proteins. However, the effectiveness of such algorithms is currently limited by some fundamental factors, including (1) the low a-priori probability of novel proteins participating in a detailed function; (2) the huge false data present in high-throughput datasets; (3) the incomplete data coverage of functional classes; (4) the abundant but heterogeneous negative samples for training the algorithms; and (5) the lack of detailed functional knowledge for training algorithms. Here, for partially characterized proteins, we suggest an approach to finding their finer functions based on protein interaction sub-networks or gene expression patterns, defined in function-specific subspaces. The proposed approach can lessen the above-mentioned problems by properly defining the prediction range and functionally filtering the noisy data, and thus can efficiently find proteins’ novel functions. For thousands of yeast and human proteins partially characterized, it is able to reliably find their finer functions (e.g., the translational functions) with more than 90% precision. The predicted finer functions are highly valuable both for guiding the follow-up wet-lab validation and for providing the necessary data for training algorithms to learn other proteins.     

The controlling factors and distribution prediction of H2S formation in marine carbonate gas reservoir, China

Generally, there are some anhydrites in carbonate reservoir, as H2S is also familiar in carbonate oil and gas reservoirs. Nowadays, natural gas with high H2S concentration is usually considered as TSR origin, so there is close relationship between H2S and anhydrite. On the contrary, some carbonate rocks with anhydrite do not contain H2S. Recently, researches show that H2S is only a necessary condition of H2S formation. The reservoir porosity, sulfate ion content within formation water, reservoir temperature, oil/gas and water interface, hydrocarbon and some elements of reservoir rock have great controlling effects on the TSR occurrence. TSR deoxidizes hydrocarbon into the acidic gas such as H2S and CO2, and the H2S formation is controlled by TSR occurrence, so the relationship among reaction room, the contact chance of sulfate ion and hydrocarbon, the reservoir temperature has great influence on the TSR reaction. H2S has relatively active chemical quality, so it is still controlled by the content of heavy metal ion. Good conditions of TSR reaction and H2S preservation are the prerequisite of H2S distribu- tion prediction. This paper builds a predictive model based on the characteristic of natural gas reservoir with high H2S-bearing. In the porosity reservoir with anhydrite, the formation water is rich in sulfate and poor in heavy metal ion. Oil and gas fill and accumulate in the gas reservoir with good preservation conditions, and they suffered high temperature later, which indicates the profitable area of natural gas with high H2S-bearing.     

神经网络在可靠性分析中的应用

应用人工神经网络对舰炮发生故障的发射炮弹数进行了预测。借助前馈神经网络对非线性函数的逼近能力，对发生故障的发射炮弹数进行了单步预测，并与使用灰色数列预测的结果进行了对比，验证了人工神经网络进行可靠性预测的可行性。