半监督学习方法研究综述

半监督学习存在于现实世界的各个场景中，可在生物化学领域对科学研究产生巨大的作用。在各领域也都有相关具体应用，如病毒毒性预测、网络安全检测、软传感器的应用等。随着机器学习领域的不断突破，目前尚缺乏关于半监督学习方法研究的完整综述。本文首先给出半监督学习的定义并分析了该领域应用过程中存在的挑战；然后梳理分析了半监督学习的四种方法，包括：半监督聚类、降维、回归、分类，并列出了这四种不同方法中比较先进的算法。随后介绍了各算法常见的评价指标（如精确率、召回率和ROC曲线等），对比了各类半监督学习算法效果，研究发现半监督学习方法都存在高于完全监督学习支持向量机的准确率，其中SSC-EKE算法以绝对优势领先传统的支持向量机经典监督学习算法。最后介绍了半监督学习的实际应用场景，展望了半监督学习的未来研究方向，并对全文进行总结。     

基于PCNN捕获-抑制机制的图像分类器设计

针对医学图像识别的重要性,本文提出了一种结合彩色图像特征场和基于脉冲耦合神经网络捕获-抑制机制的图像分类器设计方法.首先,构造符合人类视觉的彩色图像特征场;然后,利用脉冲耦合神经网络设计分类器实现对图像的识别.最后,应用于JAFFE人脸表情数据库和中医舌象的分类识别中,并与ADABOOST方法和SVM方法进行比较.实验结果表明,本文方法可以获得更好的医学图像识别效果.     

手写汉字的并行紧致集成识别方法

针对手写汉字识别这一大类别分类问题, 给出了一种新的基于多层感知机(MLP)网络的并行紧致集成方法. 方法中强调综合集成中人机结合的观点, 编码训练紧致网络分类器, 并通过一个两步监督学习过程实现整个集成网络分类系统. 与以前的集成方法相比较, 并行紧致集成方法在结构和信息处理方式上具有并行紧致的特点, 体现出较高的分类性能, 为MLP网络在大类别分类问题中的应用提供了一条可行的途径.     

支持向量机及其在机械故障诊断中的应用研究

针对支持向量机分类算法中模型选择对分类精确性影响很大的问题，结合转子实验台模拟的典型旋转机械故障数据对影响多故障分类器分类性能的相关因素进行了研究。结果表明，在少量时域故障数据样本条件下，选用不同的核函数及核函数参数对多故障分类器的分类精度有一定影响，为实际工程应用中选择合适的支持向量机核函数类型及其参数提供一定的帮助.     

一种具有记忆自学习能力的快速动态寻优算法及其无功优化求解

针对传统人工智能在随机复杂环境的适应及交互能力较低问题,有机地将经典强化学习Q(?)算法与多主体协同行为进行高度融合,提出了一种具有记忆自学习能力的快速动态寻优算法.该算法通过与外部环境反复的交互来进行自学习改进,并利用值函数矩阵储存状态-动作对记忆,提出了联系记忆方式,有效地对传统Q(?)算法的动作空间进行降维处理,减小了记忆矩阵的规模;基于多主体协同合作的概念,采用多个主体同时对记忆矩阵进行迭代更新,明显提高了更新速度;在预学习形成良好的记忆后,能快速地进行在线动态优化.最后,文章利用电力系统经典无功优化模型进行了算法测试,IEEE 118节点和IEEE 300节点标准算例仿真表明:本文所提算法在保证较高收敛性的同时,寻优速度能提高到遗传算法、蚁群算法、粒子群等传统人工智能方法的5~40倍,非常适用于大规模复杂电网的在线滚动无功优化.     

C/C复合材料CVI工艺人工神经网络建模

C/C复合材料CVI制备工艺过程的本质繁杂性限制了该材料的广泛应用. 尝试利用人工神经网络技术对该工艺过程进行辨识与仿真, 采用Levenberg-Marquardt算法建立了通用的CVI工艺神经网络模型. 根据CVI工艺复杂、参数众多等特点, 结合有限元技术及工艺实验从教师样本处理、网络拓扑结构设计和学习参数调整等方面对网络学习算法作了进一步的改进. 通过对等温CVI样本集的学习, 初步建立了管类零件等温CVI工艺知识库. 结果表明: 该模型可以挖掘样本蕴含的领域知识, 不仅可以对单个工艺参数的时间效应进行预测和分析, 而且可以分析任意两个工艺参数对致密化过程的偶合作用.     

BP网络学习能力与泛化能力满足的不确定关系式

分析BP网络过拟合出现时网络学习能力及泛化能力与其他影响因素之间的内在联系, 引入复相关系数描述样本复杂性程度; 遵循计算不确定性原理和神经网络结构设计的最简原则, 类比信息传递过程中的一般测不准关系式, 建立了BP网络过拟合出现时, 反映网络学习能力的训练样本集的训练相对误差与表征泛化能力的网络对检验样本集的测试相对误差之间满足的不确定关系式; 通过模拟多种不同类型函数的BP网络过拟合数值模拟实验, 确定了关系式中过拟合参数q的取值范围一般为7×10^−3~7×10^−2; 依据不确定关系式, 导出了在用复相关系数描述样本复杂性和满足给定逼近误差要求下, 网络具有较佳泛化能力的隐节点数的计算公式, 并验证了其合理性; 指出BP网络应用于给定样本集的训练过程中, 为改进泛化能力的训练最佳停止方法.     

基于样本密度的非监督动态改进FCM算法

针对传统FCM算法中聚类结果对初始聚类中心敏感,分类数C依赖于先验知识确定的不足,提出了基于样本密度的非监督动态改进FCM算法。该算法基于初始样本密度确定初始聚类中心,基于模糊伪F统计量自动确定最佳分类数,实现了基于样本密度非监督状态下的动态聚类。选取瓦斯涌出量的聚类分析进行应用,应用结果表明,该算法能合理选择初始聚类中心并动态确定分类数,降低了对初始聚类中心的依赖度,提高了收敛速度和自动化程度,并能根据指标做出正确预测。     

基于支持向量机和粒子群算法的电子鼻伤口感染检测

针对传统的伤口感染诊断方法耗时长,操作复杂等问题,提出了一种基于电子鼻和支持向量机(SVM)的方法进行伤口感染检测,分别检测非感染和三种常见病原菌感染的大白鼠伤口顶空气体,然后利用 SVM对实验数据进行识别.同时,鉴于传感器阵列的优化以及 SVM参数选择对其分类准确率有重大的影响,提出一种基于粒子群算法(PSO)的传感器阵列和 SVM参数同步优化方法.实验结果表明,SVM结合 PSO与传统的神经网络以及遗传算法相比,极大提高伤口感染检测的准确率     

基于Boosting梯度下降理论的时间序列建模方法

在预测问题中有很多实际问题具有时间序列特性,如何分析并建立合理的时间序列预测模型具有重要的理论价值与广泛的应用前景.随着对时间序列研究的逐步深入,时间序列预测模型变得越来越复杂,建模过程对设计技巧依赖性越来越强.如何设计一种不依赖于操作者的能力和经验,并且具有良好效果的建模方法呢？本文在集成学习理论的基础上,提出了基于Boosting梯度下降理论的自回归模型集成（AREnsembleLearning,AREL）建模法,从理论上分析并论证了该建模法的有效性.为了避免AREL精确拟合每一个训练样本点,在建模过程中引入了？-不敏感损失函,从而降低随机噪声对时间序列预测模型的影响.进而为了提高AREL对非平稳时间序列的处理能力,提高算法的鲁棒性,防止发生过拟合,降低算法对模型阶次设置的敏感性,提出了基于AREL的加权k近邻（weightedkNN）预报法.通过实例测试,并对结果进行了讨论,验证了所提出的建模法与预报方法的有效性.     

A Meta‐learning Framework for Bankruptcy Prediction

The implication of corporate bankruptcy prediction is important to financial institutions when making lending decisions. In related studies, many bankruptcy prediction models have been developed based on some machine‐learning techniques. This paper presents a meta‐learning framework, which is composed of two‐level classifiers for bankruptcy prediction. The first‐level multiple classifiers perform the data reduction task by filtering out unrepresentative training data. Then, the outputs of the first‐level classifiers are utilized to create the second‐level single (meta) classifier. The experiments are based on five related datasets and the results show that the proposed meta‐learning framework provides higher prediction accuracy rates and lower type I/II errors when compared with the stacked generalization classifier and other three widely developed baselines, such as neural networks, decision trees, and logistic regression. Copyright © 2011 John Wiley & Sons, Ltd.     

机器学习面临的挑战

该文讨论了机器学习目前面临的几个挑战，包括：高维特征空间和数据量问题，大数据量的计算困难，寻求最优解的困难和可解释性差等问题．然后针对当前很多人关心的几个重要问题，例如大数据问题，深度学习，概率图模型等做了分析，以引起深入思考．     

Incorporating textual and management factors into financial distress prediction: A comparative study of machine learning methods

Financial distress prediction (FDP) has been widely considered as a promising approach to reducing financial losses. While financial information comprises the traditional factors involved in FDP, nonfinancial factors have also been examined in recent studies. In light of this, the purpose of this study is to explore the integrated factors and multiple models that can improve the predictive performance of FDP models. This study proposes an FDP framework to reveal the financial distress features of listed Chinese companies, incorporating financial, management, and textual factors, and evaluating the prediction performance of multiple models in different time spans. To develop this framework, this study employs the wrapper-based feature selection method to extract valuable features, and then constructs multiple single classifiers, ensemble classifiers, and deep learning models in order to predict financial distress. The experiment results indicate that management and textual factors can supplement traditional financial factors in FDP, especially textual ones. This study also discovers that integrated factors collected 4 years prior to the predicted benchmark year enable a more accurate prediction, and the ensemble classifiers and deep learning models developed can achieve satisfactory FDP performance. This study makes a novel contribution as it expands the predictive factors of financial distress and provides new findings that can have important implications for providing early warning signals of financial risk.     

A deep residual compensation extreme learning machine and applications

The extreme learning machine (ELM) is a type of machine learning algorithm for training a single hidden layer feedforward neural network. Randomly initializing the weight between the input layer and the hidden layer and the threshold of each hidden layer neuron, the weight matrix of the hidden layer can be calculated by the least squares method. The efficient learning ability in ELM makes it widely applicable in classification, regression, and more. However, owing to some unutilized information in the residual, there are relatively huge prediction errors involving ELM. In this paper, a deep residual compensation extreme learning machine model (DRC-ELM) of multilayer structures applied to regression is presented. The first layer is the basic ELM layer, which helps in obtaining an approximation of the objective function by learning the characteristics of the sample. The other layers are the residual compensation layers in which the learned residual is corrected layer by layer to the predicted value obtained in the previous layer by constructing a feature mapping between the input layer and the output of the upper layer. This model is applied to two practical problems: gold price forecasting and airfoil self-noise prediction. We used the DRC-ELM with 50, 100, and 200 residual compensation layers respectively for experiments, which show that DRC-ELM does better in generalization and robustness than classical ELM, improved ELM models such as GA-RELM and OS-ELM, and other traditional machine learning algorithms such as support vector machine (SVM) and back-propagation neural network (BPNN).     

Judgement in learning‐curve forecasting: a laboratory study

This study investigates whether human judgement can be of value to users of industrial learning curves, either alone or in conjunction with statistical models. In a laboratory setting, it compares the forecast accuracy of a statistical model and judgemental forecasts, contingent on three factors: the amount of data available prior to forecasting, the forecasting horizon, and the availability of a decision aid (projections from a fitted learning curve). The results indicate that human judgement was better than the curve forecasts overall. Despite their lack of field experience with learning curve use, 52 of the 79 subjects outperformed the curve on the set of 120 forecasts, based on mean absolute percentage error. Human performance was statistically superior to the model when few data points were available and when forecasting further into the future. These results indicate substantial potential for human judgement to improve predictive accuracy in the industrial learning‐curve context. Copyright © 1999 John Wiley & Sons, Ltd.     

Using the yield curve to forecast economic growth

This paper finds the yield curve to have a well-performing ability to forecast the real gross domestic product growth in the USA, compared to professional forecasters and time series models. Past studies have different arguments concerning growth lags, structural breaks, and ultimately the ability of the yield curve to forecast economic growth. This paper finds such results to be dependent on the estimation and forecasting techniques employed. By allowing various interest rates to act as explanatory variables and various window sizes for the out-of-sample forecasts, significant forecasts from many window sizes can be found. These seemingly good forecasts may face issues, including persistent forecasting errors. However, by using statistical learning algorithms, such issues can be cured to some extent. The overall result suggests, by scientifically deciding the window sizes, interest rate data, and learning algorithms, many outperforming forecasts can be produced for all lags from one quarter to 3 years, although some may be worse than the others due to the irreducible noise of the data.     

Learning dynamics in the formation of European inflation expectations

We investigate learning dynamics in the formation of household inflation expectations in the six largest euro area countries. Our findings reveal heterogeneity in the learning rules that European households use to forecast inflation. We also find pronounced heterogeneity in the way consumers process new data. These differences vary not only across countries but also over time, suggesting that the learning behavior of households is state dependent.     

星形胶质细胞在学习记忆中的作用

星形胶质细胞是神经胶质细胞的主要组成部分,遍布整个神经系统.以往认为星形胶质细胞的功能主要为支持营养神经元,参与物质代谢,血脑屏障形成等等,但最近研究发现它们还可促进突触形成,影响信号传导及突触可塑性,从而在学习记忆的形成中发挥重要作用.     

A new PM2.5 concentration forecasting system based on AdaBoost-ensemble system with deep learning approach

A reliable and efficient forecasting system can be used to warn the general public against the increasing PM_2.5 concentration. This paper proposes a novel AdaBoost-ensemble technique based on a hybrid data preprocessing-analysis strategy, with the following contributions: (i) a new decomposition strategy is proposed based on the hybrid data preprocessing-analysis strategy, which combines the merits of two popular decomposition algorithms and has been proven to be a promising decomposition strategy; (ii) the long short-term memory (LSTM), as a powerful deep learning forecasting algorithm, is applied to individually forecast the decomposed components, which can effectively capture the long-short patterns of complex time series; and (iii) a novel AdaBoost-LSTM ensemble technique is then developed to integrate the individual forecasting results into the final forecasting results, which provides significant improvement to the forecasting performance. To evaluate the proposed model, a comprehensive and scientific assessment system with several evaluation criteria, comparison models, and experiments is designed. The experimental results indicate that our developed hybrid model considerably surpasses the compared models in terms of forecasting precision and statistical testing and that its excellent forecasting performance can guide in developing effective control measures to decrease environmental contamination and prevent the health issues caused by a high PM_2.5 concentration.