Using Informative Score for Instance Selection Strategy in Semi-Supervised Sentiment Classification

Sentiment classification is a useful tool to classify reviews about sentiments and attitudes towards a product or service.Existing studies heavily rely on sentiment classification methods that require fully annotated inputs.However,there is limited labelled text available,making the acquirement process of the fully annotated input costly and labour-intensive.Lately,semi-supervised methods emerge as they require only partially labelled input but perform comparably to supervised methods.Nevertheless,some works reported that the performance of the semi-supervised model degraded after adding unlabelled instances into training.Literature also shows that not all unlabelled instances are equally useful;thus identifying the informative unlabelled instances is beneficial in training a semi-supervised model.To achieve this,an informative score is proposed and incorporated into semisupervised sentiment classification.The evaluation is performed on a semisupervised method without an informative score and with an informative score.By using the informative score in the instance selection strategy to identify informative unlabelled instances,semi-supervised models perform better compared to models that do not incorporate informative scores into their training.Although the performance of semi-supervised models incorporated with an informative score is not able to surpass the supervised models,the results are still found promising as the differences in performance are subtle with a small difference of 2%to 5%,but the number of labelled instances used is greatly reduced from100%to 40%.The best finding of the proposed instance selection strategy is achieved when incorporating an informative score with a baseline confidence score at a 0.5:0.5 ratio using only 40%labelled data.

Public Speaking的技巧与策略(8)——如何进行Informative Presentation

演讲和面对面谈话非常相似。如果我们想教授某一种技能,如演示性演讲,我们必须首先让听众知道我们想让他们了解什么,做什么。如果是劝说式演讲或辩论,我们必须在劝服听众相信我们的观点前使他们了解所要讨论的题目。这种告知听众的做法要求我们清晰简洁地提出我们的主题,然后用具体的细节、例证来支持主题或在辩论中提出具体明确的参考材料。

Old nest material functions as an informative cue in making nest-site selection decisions in the European Kestrel(Falco tinnunculus)

Background:For secondary cavity-nesting bird species that do not add lining materials to nests,the presence of old nest material or organic remains that have accumulated within nest cavities from previous breeding events may be a cue of nest-site quality.These materials potentially contain information about past breeding success in con-and heterospecifics and may improve the thermal insulation of eggs during incubation.However,few studies have addressed whether the presence of old nest materials serves as a cue for cavity-nesting raptors when choosing specific nest sites.Methods:We conducted a 9-year nest box experiment to test whether old nest materials from con-and heterospecifics serve as informative cues to the European Kestrel(Falco tinnunculus)when making nest selection decisions,as this species uses nest boxes without adding nesting material.Results:The presence of old nest materials and entrance size best discriminated nest boxes occupied by European Kestrels from unoccupied boxes.Nest boxes containing conspecific organic remains,artificial dry leaf and branch material,and material left behind by Great Tits(Parus major)were reused at higher rates,especially those containing conspecific nest material,than nest boxes containing true or simulated nest materials from predators.In 2010,no single nest box was occupied by the same banded individual that occupied the box in the previous year(10 females and 2 males were banded in 2009).Conclusions:European Kestrels preferred nest boxes containing old nest material over empty boxes,which is consistent with previous findings that they exploit con-and heterospecific cues when deciding where to settle and breed,as old nest or organic material provides substrate for incubating females.Kestrels may be able to assess the predation risks associated with a specific nest site based on experience or the presence of prey remains.The repeated use of nest boxes across breeding seasons by kestrels cannot be entirely ascribed to philopatry.This study provides evidence that old nest materials are potentially used as informative cues when making nest-site selection decisions in European Kestrels.

The Fourier Notation of the Geomagnetic Signals Informative Parameters

The paper discusses the quantitative definition of the s/n (signal to noise ratio) by means of new computational parameters derived (and computed) by the Fourier analysis. The theme is of great relevance when the geomagnetic observed field has high transient noise and high energy content (i.e.geomagnetic signal interfered by human activity magnetic band) and when the signal analysis action is oriented to the detection of magnetic sources characterized by quasi-punctiform size, low energy level and kinetic mechanical status (i.e.uw armed terrorist). The paper shows the results obtained introducing two new informative spectral parameters: the informative capability “C” and the enhanced informative capability “eC”. These parameters are depending on the comparison of the energy of the target signal with total field energy and they are characteristics of each elementary signal. C classifies the energy of the spectrum in two metrological bands: elementary signal informative energy EI (band or single signal) and passive energy EP. This metrological classification of the energy overtakes the concept of noise: each signal is part of the noise band when it is not under observation and becomes out of the band when it is under observation (numerical observation→computation). C (and eC) allows to compute the value of the “visibility” of the informative signals in a high energy geomagnetic field (or spectrum). C is a fundamental parameter for the evaluation of the effectiveness of singularity magnetic metrology in the passive detection of small magnetic sources in high noised magnetic field.

Informative Property of the Data Set in a Single-input Single-output (SISO) Closed-loop System with a Switching Controller

Closed-loop identification is important and necessary to various model-based advanced process control strategies, whose performance depends greatly on the informative property of the data set. Switching control is an important method in process control. Therefore, this paper studies the informative property of a data set in a single-input single-output (SISO) closed-loop system with a switching controller. It is proved that this data set is informative if the controller switches among at least two modes (i.e., feedback laws). Our result does not require any assumption on the way of switch and removes the constraints on the switching manner required in some classical literature. Finally, simulation case studies based on a continuous stirred-tank reactor (CSTR) process are given to validate the results.

Semiparametric Estimator of Mean Conditional Residual Life Function under Informative Random Censoring from Both Sides

In this paper we study estimator of mean residual life function in fixed design regression model when life times are subjected to informative random censoring from both sides. We prove an asymptotic normality of estimators.

Quality Requirements for Multimedia Interactive Informative Systems

Multimedia Interactive Informative Systems (MIIS) are software applications resulting from the convergence of multiples technologies such as audiovisual, computing and communication. They aim to transmit information to a large, diverse and dispersed public. As with other mass media, the fulfillment of MIIS goals depends largely on the quality of communication between the system and end users. Therefore, those systems should be developed in order to ensure that this quality requirement is satisfied. If MIIS are constructed according to usual software engineering practices, the analysis discipline of the development process includes requirements identification and specification;however, these techniques are focused on functional requirements, and they do not give much importance to non-functional requirements. In this paper, we propose a development process based on the production of videogames which has two different phases: preproduction and production. The first phase, corresponding to requirements identification, derives into the concept of system. In order to translate this concept into a specification, we propose the use of new communicational attributes and a MIIS metamodel. The establishment of MIIS non-functional specification is the result of analyzing class diagrams through quality attributes. In order to evaluate if the specifications are responding to communicational attributes, a functional prototype is built and evaluated with end users. The proposed methodology is applied to a real case study.

On the Informative Function of Language in the Poems and Couplets in Benzhu Temples in Dali,Yunnan

In Dali Bai Autonomous Prefecture,Yunnan Province,there exists a quite special and interesting cultural phenomenon of the Bai ethnic group,which is the belief in Benzhu.There are a lot of poems and couplets in Benzhu temples.The poems are carved on the walls and tablets.The couplets,being on both sides of the doors,are carved on planks or written on red paper.Lan guage used in the poems and couplets plays the informative function.

Intertextile Beijing Features More Informative programme of seminars and trend forums add value to fair

Intertextile Beijing Apparel Fabrics,will be held from 29-31 March 2009 at the China International Exhibition Centre,will showcase the latest textiles from around the world on 48,000 sqm of exhibition space.The event has confirmed 1100 exhibitors from 14 countries and regions including

Weighted-averaging-based classification of laser-induced breakdown spectroscopy measurements using most informative spectral lines

In this study,efficient spectral line selection and wcightcd-avcraging-bascd processing schemes are proposed for the classification of laser-induced breakdown spectroscopy(UBS)measurements.For fast on-line classification,a set of representative spectral lines arc selected ami processed relying on the information metric,instead of the time consuming full spectrum based analysis.I he most informative spectral line sets arc investigated by the joint mutual information estimation(MIR)evaluated with the Gaussian kernel density,where dominant intensity peaks associated with the concentrated components arc not necessarily most valuable for classification.In order to further distinguish the characteristic patterns of die LIBS measured spectrum,two-dimensional spectral images are synthesized through column-wise concatenation of the peaks along with their neighbors.For fast classification while preserv ing die effect of distinctive peak patterns,column-wise Gaussian weighted averaging is applied to die synthesized images,yielding a favorable trade off between classification performance and computational complexity.To explore the applicability of the proposed schemes,two applications of alloy classification and skin cancer detection arc investigated with the multi-class and binary support vector machines classifiers,respectively.Ihc MIE measures associated with selected spectral lines in bodi applications show a strong correlation to the actual classification or detection accuracy,which enables to find out meaningful combinations of spectral lines.In addition,the peak patterns of the selected lines and their Gaussian weighted averaging with nciehbors of the selected peaks efficiently distineuish different classes of LIBS measured spectrum.

Research of Impact of Geografical Latitute and Residual Ionospheric Noises on Informativeness of Measuring of Zenith Wet Delay of GPS Signals

It is noted that necessity of further increase of accuracy of GPS positioning systems requires de-velopment of more perfect methods to compensate information losses occurred due to residual ionospheric delay by using optimization procedures. According to the conditions of formulated optimization task, the signal/noise ratio in measurements of zenith wet delay depends on the second order ionospheric errors, geographic latitude and day of year. At the same time if we assume that the number of measurements at the fixed geographic site is proportional to geographic latitude and if we accept existence of only two antiphase scenarios for variation of residual ionospheric delay on latitude normed by their specific constant, there should be optimum functional dependence of precipitated water on latitude upon which the quantity of measuring information reaches the maximum. The mathematical grounding of solution of formulated optimization task is given.

Employee Stock Ownership Plans and Stock-price Informativeness

This study examines the impact of employee stock ownership plans(ESOPs)on stock-price informativeness in Chinese stock markets.Its findings indicate that firms implementing ESOPs experienced an average 11.89 percent increase in stock-price informativeness.The plans improved stock-price informativeness through increased external attention and supervision.An event study shows that ESOPs gave rise to an announcement effect,driven by anticipated performance improvements and the novelty associated with ESOPs.A mechanism analysis demonstrates that the implementation of ESOPs attracted market attention,and the increased market supervision resulting from this mitigated the moral hazards of management associated with ESOPs.Plans with more positive signals exerted a greater influence.Notably,ESOPs that prioritized management incentives gained more recognition in the market.As the incentive effects of ESOPs were weaker than those of equity incentive plans and the ESOPs lost novelty over time,the annual announcement effect diminished gradually.These findings underscore the necessity of strengthening ESOP incentives for continued optimization of priceefficiency.

Regression Analysis of Misclassified Current Status Data with Informative Observation Times

Misclassified current status data arises if each study subject can only be observed once and the observation status is determined by a diagnostic test with imperfect sensitivity and specificity.For the situation,another issue that may occur is that the observation time may be correlated with the interested failure time,which is often referred to as informative censoring or observation times.It is well-known that in the presence of informative censoring,the analysis that ignores it could yield biased or even misleading results.In this paper,the authors consider such data and propose a frailty-based inference procedure.In particular,an EM algorithm based on Poisson latent variables is developed and the asymptotic properties of the resulting estimators are established.The numerical results show that the proposed method works well in practice and an application to a set of real data is provided.

A New Approach for Regression Analysis of Multivariate Current Status Data with Informative Censoring

Regression analysis of interval-censored failure time data has recently attracted a great deal of attention partly due to their increasing occurrences in many fields.In this paper,we discuss a type of such data,multivariate current status data,where in addition to the complex interval data structure,one also faces dependent or informative censoring.For inference,a sieve maximum likelihood estimation procedure is developed and the proposed estimators of regression parameters are shown to be asymptotically consistent and efficient.For the implementation of the method,an EM algorithm is provided,and the results from an extensive simulation study demonstrate the validity and good performance of the proposed inference procedure.For an illustration,the proposed approach is applied to a tumorigenicity experiment.

高速铁路日常客运量的EMD-Informer组合预测方法

铁路客流需求的科学预测是进行运输组织方案决策的重要依据。以高速铁路历史客票数据为基础,结合经验模态分解(empirical mode decomposition,EMD)与机器深度学习中的注意力机制,提出高速铁路日客流量的EMD-Informer组合预测方法。首先采用EMD方法分解高速铁路客流量序列,获得具有周期特征和线路客流内在特征的模态分量,再利用Informer模型分别训练和预测各模态分解分量,并通过多头注意力机制高效挖掘客流数据内在规律和捕捉数据序列中的关键特征,在此基础上重组各分量预测值,从而得到高速铁路日常客流的整体高精度预测值。同时,根据结合问题特征的大量实验,明确可供实际运用参考的超参数设置规则。基于京沪高速铁路全线的实例计算分析表明,相对对比预测方法,EMD-Informer组合预测方法在高速铁路客流的单步预测及超前预测上均具有明显更小的预测误差。

机场能见度临近预测方法

能见度是保障机场航班安全、正常运行的重要标准之一。为精准预测能见度,使用2020年天津机场气象和常规空气质量监测数据,构建基于方差膨胀因子(Variance Inflation Factor,VIF)、主成分分析(Principal Components Analysis,PCA)和Informer的能见度预测模型,并将均方根误差、平均绝对误差、平均绝对百分比误差作为评价指标进行误差分析。结果显示,VIF PCA Informer模型比单一的Informer和简单组合模型效果更优,能更好地捕捉长时间序列特征的关系。相比于单一的Informer、长短期记忆神经网络和门控循环单元模型,VIF PCA Informer模型均方根误差下降了0.2141~0.3486,平均绝对误差下降了0.1842~0.2753,平均绝对百分比误差下降了0.3224~0.5270;VIF PCA Informer模型对能见度的临近预测(1 h)更为精准。使用高效的机场能见度预测模型可在保障航班安全高效运行方面发挥较大支撑作用。

基于时间序列的发电机设备异常分析

为提高发电机组设备运行维护管理水平,提出一种基于PCA-Informer方法的发电设备故障预测技术。首先使用主成分分析(PCA)算法降低时间序列数据的特征维度;其次将数据分批次输入Encoder中,由Encoder执行蒸馏操作,提取长时间序列输入间的Long-Range依赖,通过蒸馏操作为重要特征赋予更高的权重,并在下一层生成聚焦的Self-Attention Feature Map;最后由Decoder通过一个正向过程一步生成长序列输出。通过实验验证,该方法能够有效地对发电设备的故障进行预测。

寒冷条件下高拱坝施工过程仿真方法研究

众多高拱坝施工面临着恶劣寒冷气候的考验。然而,现有的高拱坝施工过程仿真方法较少直接考虑气温对施工的影响,且未考虑基于气温预测成果进行动态分块,以分析寒冷条件下利用正温时段进行浇筑的特点。针对以上问题,本文提出了基于Informer的寒冷条件下高拱坝施工过程仿真方法。首先,提出基于Informer气温预测模型,实现未来气温序列的预测,在此基础上实现可施工时段分析;其次,建立考虑动态分块的寒冷条件下高拱坝施工进度仿真模型,提出更加贴近实际施工情况的分块施工仿真策略,以更好地模拟和分析不同气温条件下的施工过程;最后,以西南地区的叶巴滩高拱坝工程为例进行研究,采用Informer模型对气温进行预测,平均误差为±1.49℃,每日的可施工时长平均误差为±1.16 h。通过对比三种不同的仿真策略,发现寒冷条件下采用动态分块策略可以充分利用可浇筑时段,施工效率更高,也表明在仿真中“考虑气温”比简单降效处理更加贴近实际情况。

考虑多时间尺度信息的风力发电机滚动轴承故障预测

风电机组滚动轴承故障会造成风电机组长时间停机,为准确预测风电机组滚动轴承故障,提出一种考虑多时间尺度信息的风力发电机滚动轴承故障预测方法。首先,采用连续变分模式分解(successive variational mode decomposition,SVMD)自适应提取轴承健康数据温度多维特征;其次,将分解的本征模态函数(intrinsic mode functions,IMFs)输入Informer模型提取多尺度时间信息训练,基于树状结构Parzen密度估计的非标准贝叶斯优化算法(tree structure Parzen density estimation,TPE)优化Informer模型超参数;然后,构建基于残差的故障指标,采用核密度估计(kernel density estimation,KDE)确定故障预警阈值;最后,将运行数据输入训练后的Informer模型进行故障预测。选取某风电场的风力发电机轴承温度数据进行故障预测,仿真结果表明,考虑多时间尺度信息的SVMD-TPE-Informer模型在发电机轴承温度预测上具有更高的预测精度和计算效率,所提方法在两个故障案例上分别能够提前15.5 h和10 h预测到故障,且不会出现误报现象,验证所提模型的有效性和稳定性。

基于STL-Informer-BiLSTM-XGB模型的供热负荷预测

供热负荷预测是指导供热系统调控的重要手段。提高供热负荷预测精度十分重要,针对机器学习中输出目标的分解预测,提出了一种基于季节和趋势分解(seasonal and trend decomposition using loess,STL)的供热负荷预测方法,构建了适用于供热负荷预测的输出目标。首先利用STL算法将供热负荷时间序列数据分解为趋势分量、周期分量和残差分量,分别训练Informer、BiLSTM和XGB模型,将构建好的3个分量预测模型的输出叠加作为初步预测结果,分析误差序列,以BiLSTM预测误差提高模型精度,构建出STL-Informer-BiLSTM-XGB预测模型。将上述模型与常用预测模型进行对比,结果表明所构建的STL-Informer-BiLSTM-XGB模型的MAPE、MAE和MSE分别为0.871%、96.18和13202.2,预测效果最优,验证了所提出的方法具有较高的供热负荷预测精度。