关键词： 地下水位   高斯过程   机器学习  

摘要： 高斯过程是近年来发展起来的一种新的机器学习方法,它有着严格的统计学习理论基础,对处理高维数非线性小样本复杂问题具有良好的适应性。针对地下水位的动态变化是一个复杂的非线性过程,地下水位与其影响因素之间存在着复杂的非线性关系,将高斯过程机器学习方法应用于地下水位预测,并提出相应的模型。工程实例研究表明,地下水位预测的高斯过程机器学习模型是科学可行的,预测精度高,简单实用,对地下水位预测的问题具有较好的适用性。

关键词： Random Forest   Nuclear Magnetic Resonance Spectrum   Machine Learning Method   Numerical Attribute   Variable Importance Measure  

摘要： Background: Current efforts in Metabolomics, such as the Human Metabolome Project, collect structures of biological metabolites as well as data for their characterisation, such as spectra for identification of substances and measurements of their concentration. Still, only a fraction of existing metabolites and their spectral fingerprints are known. Computer- Assisted Structure Elucidation (CASE) of biological metabolites will be an important tool to leverage this lack of knowledge. Indispensable for CASE are modules to predict spectra for hypothetical structures. This paper evaluates different statistical and machine learning methods to perform predictions of proton NMR spectra based on data from our open database NMRShiftDB. Results: A mean absolute error of 0.18 ppm was achieved for the prediction of proton NMR shifts ranging from 0 to 11 ppm. Random forest, J48 decision tree and support vector machines achieved similar overall errors. HOSE codes being a notably simple method achieved a comparatively good result of 0.17 ppm mean absolute error. Conclusion: NMR prediction methods applied in the course of this work delivered precise predictions which can serve as a building block for Computer-Assisted Structure Elucidation for biological metabolites.

关键词： Aging   Classification   Longevity   Shrunken centroid   T-cell subset   Weight  

摘要： Prediction of individual life span based on characteristics evaluated at middle-age represents a challenging objective for aging research. In this study, we used machine learning algorithms to construct models that predict life span in a stock of genetically heterogeneous mice. Life-span prediction accuracy of 22 algorithms was evaluated using a cross-validation approach, in which models were trained and tested with distinct subsets of data. Using a combination of body weight and T-cell subset measures evaluated before 2 years of age, we show that the life-span quartile to which an individual mouse belongs can be predicted with an accuracy of 35.3% (+/- 0.10%). This result provides a new benchmark for the development of life-span-predictive models, but improvement can be expected through identification of new predictor variables and development of computational approaches. Future work in this direction can provide tools for aging research and will shed light on associations between phenotypic traits and longevity.

关键词： knowledge discovery   machine learning method   otoneurology  

摘要： We have been interested in developing an otoneurological decision support system that supports diagnostics of vertigo diseases. in this study, we concentrate on testing its inference mechanism and knowledge discovery method. Knowledge is presented as patterns of classes. Each pattern includes attributes with weight and fitness values concerning the class. With the knowledge discovery method it is possible to form fitness values from data. Knowledge formation is based on frequency distributions of attributes. Knowledge formed by the knowledge discovery method is tested with two vertigo data sets and compared to experts' knowledge. The experts' and machine learnt knowledge are also combined in various ways in order to examine effects of weights on classification accuracy. The classification accuracy of knowledge discovery method is compared to 1- and 5-nearest neighbour method and Naive-Bayes classifier. The results showed that knowledge bases combining machine learnt knowledge with the experts' knowledge yielded the best classification accuracies. Further, attribute weighting had an important effect on the classification capability of the system. When considering different diseases in the used data sets, the performance of the knowledge discovery method and the inference method is comparable to other methods employed in this study. (C) 2008 Elsevier Ireland Ltd. All rights reserved.

关键词： hERG potassium channel inhibitor   Machine learning methods   Molecular descriptors   Recursive feature elimination (RFE)   Support vector machine(SVM)  

摘要： The human Ether-a-Go-Go-Related Gene (hERG) potassium ion channel plays a crucial role in cardiac repolarization. The inhibition of this channel by marketed drug may prolong the length of time between the start of Q wave and end of the T wave oil an electrocardiogram (QT interval) and then possibly leads to fatal cardiac arrhythmia. Therefore, it is vital to predict the potential hERG inhibitors in early stages of drug discovery. This work explored the identifying of hERG inhibitors by different machine learning methods including C4.5 Decision Tree (C4.5 DT), multilayer perceptron, Radial Basis Function Network (RBFNetwork), and Support Vector Machine (SVM). Recursive Feature Elimination (RFE), a feature selection method, was used to select molecular descriptors appropriate for distinguishing hERG inhibitors and noninhibitors. For Data_91 classification system with 91 compounds., the prediction accuracies of those methods are 72.1 - 92.3% for hERG inhibitors and 81.7 - 98.1% for noninhibitors in a five-fold crossvalidation calculation. For an independent validation set of 47 compounds of Data 91, these methods gave an overall accuracy of 72.3-80.9% using the selected descriptors. This suggests that the combination of the machine learning methods and the feature selection method can provide an effective way to filter the hERG inhibitors in the drug discovery process.

关键词： 模糊c-means算法   模糊隶属度   空间自相关   空间数据挖掘   空间分类和预测  

摘要： 空间分类和预测是空间数据挖掘中一个非常重要的方法,但对它们的研究目前尚处于初始阶段.通过引入空间对象对模糊聚类的模糊隶属度的概念,提出了基于模糊c-means算法的空间数据分类和预测的方法(SFCM).该方法首先用模糊c-means方法对数据集论域空间进行聚类,但由于空间数据具有空间自相关的特性,在用模糊c-means算法进行空间聚类时加入了空间信息.然后计算每个空间对象对所有聚类的模糊隶属度并从中找出模糊隶属度最大的聚类.最后用该聚类中心对象的因变量的值作为该空间对象的因变量的估计值.理论分析和实验结果表明,该算法是有效可行的.

关键词： prediction   RF   pregnane X receptor   Received therapy or drug for  

摘要： The pregnane X receptor (PXR) regulates the expression of genes involved in xenobiotic metabolism and transport. In vitro methods to screen for PXR agonists are used widely. In the current study, computational models for human PXR activators and PXR nonactivators were developed using recursive partitioning (RP), random forest (RF), and support vector machine (SVM) algorithms with, VolSurf descriptors. Following 10-fold randomization, the models correctly predicted 82.6-98.9% of activators and 62.0-88.6% of nonactivators. The models were validated using separate test sets. The overall (n = 15) test set prediction accuracy for PXR activators with RP, RF, and SVM PXR models is 80-93.3%, representing an improvement over models previously reported. All models were tested with a second test set (n = 145), and the prediction accuracy ranged from 63 to 67% overall. These test set molecules were found to cover the same area in a principal component analysis plot as the training set, suggesting that the predictions were within the applicability domain. The FlexX docking method combined with logistic regression performed poorly in classifying this PXR test set as compared with RP, RF, and SVM but may be useful for qualitative interpretion of interactions within the LBD. From this analysis, VolSurf descriptors and machine learning methods had good classification accuracy and made reliable predictions within the model applicability domain. These methods could be used for high throughput virtual screening to assess for PXR activation, prior to in vitro testing to predict potential drug-drug interactions.

关键词： 汉语分词   机器学习   词性标注   序列标注   命名实体识别  

摘要： 词法分析是自然语言处理中的基础工作。由于汉语中分词问题的存在，使得汉语的词法分析更加困难也更加重要。近年来，基于标注语料库采用机器学习方法进行词法分析的策略取得了较大成功，但是仍然存在很多问题。\n 本文实现了基于条件随机场等序列标注机器学习方法的汉语词法分析系统，在特征选取、标记集定义、解码策略、置信度的使用等方面进行了深入研究，在对应的机器学习方法的算法和实现层面分别作了改进，在SIGHAN标准评测数据集上显示了其良好的性能。\n 系统地分析了汉语自动分词、词性标注、命名实体识别三个任务的特点和关联，提出了在机器学习框架下一体化分析的观点，实现了基于任务合并和贝叶斯推理的整体分析方法，降低了级联错误的影响，并提高了分词的性能。\n 实现了基于特征层次变换引入领域区分信息的自适应策略，解决了训练和测试中文本领域和标注规范的差异带来的数据失配问题，并讨论了同源数据中不同的领域分布对数据分布和学习算法的影响。\n 设计并实现了针对最大熵模型和条件随机场的并行训练算法，解决了大规模数据训练过程中的时间和空间代价过高的问题。分析了学习样本数量和优化算法误差对学习性能的影响，实现了基于随机梯度下降的优化算法，指出了其在大数据学习中的优越性。

关键词： 网络安全   网络异常检测   数据挖掘   机器学习  

摘要： 随着互联网的飞速发展以及网络应用服务越来越呈现出的多样化和多元化等特点，网络协议、网络基础设施以及网络服务存在的相关缺陷和问题为黑客所利用，网络攻击正趋日益增长之势，网络威胁种类也日渐增加，严重地危害了网络安全和数据安全。网络安全问题成为威胁网络资源、网络基础设施和网络应用服务的最大问题和人们的关注焦点。\n 网络异常检测技术是保证网络安全的核心技术。虽然人们对其进行了多年的研究，然而依然存在着误报率过高和检测开销大的问题，因而难以满足应用需求。基于这个背景，本文在全面研究和分析传统网络异常检测技术的基础上，深入研究了基于数据挖掘和机器学习方法的网络异常检测技术，并从网络异常检测算法研究、与网络异常检测密切相关的训练集样本选取以及特征空间处理等多个层面，进行系统化、整体化的方法性研究，从而提出了一套高效的具有较高检测性能和检测速度的网络异常检测方法。本文的主要贡献和创新表现在以下五个方面：\n ●研究、分析并综述了传统的网络异常检测技术。我们首先对网络异常检测技术的评价指标进行了介绍，并将当前的研究成果分为基于统计的检测方法、基于机器学习的检测方法、基于数据挖掘的检测方法、基于聚类的检测方法以及近年来最新出现的面向Web攻击的检测方法等几大类进行了详细地介绍和分析，包括他们的代表技术、主要优缺点以及应用现状等。其次，在此基础上，笔者对这几类异常检测技术从计算开销、检测性能、检测的特征空间维度、通用性能及适合的处理方式五个层面进行了详细的比较。最后，笔者根据当前的研究现状，展望了网络异常检测技术的未来主要研究方向。\n ●提出并实现了一种基于ETCM-KNN(Enhanced Transductive Confidence Machines forK-Nearest Neighbors)数据挖掘算法的网络异常检测新方法。针对传统的检测算法仍然存在较高误报率的不足，笔者首先将模式识别和欺诈检测领域表现卓越的TCM-KNN算法引入入侵检测领域，然后针对网络异常检测的实际应用需求对其进行改进，实现了一种基于ETCM-KNN算法的网络异常检测新方法。与传统方法仅仅依赖少量样本建立的模型进行异常判定不同，该方法采用奇异值及P值两个核心概念，通过阈值判定的方式，深入挖掘并使用正常训练数据集的整体优势，高效地使用直推式方法对网络异常进行判定，从而在很大程度上避免了传统方法在正常/异常这个二类判定问题上的不准确性。通过大量采用国际上权威的KDD Cup1999入侵检测测试数据集的测试表明，相对于该领域经典的研究成果来说，其不但具有较高的检测率(99％左右)，而且还保证了较低的误报率(2％左右)。\n ●提出并研究了使用样本选择方法对网络异常检测技术的性能优化策略。在网络异常检测技术的实际应用过程中，不可避免地需要依赖仅含“正常数据”的训练数据对正常的网络流量和网络行为进行训练和建模。然而，如果训练数据的冗余导致其规模过大，对于以数据挖掘和机器学习为基础的异常检测算法来说，必然导致较高的计算开销和内存消耗，对于本文所提出的基于距离计算的ETCM-KNN算法尤其如此。因此，我们在对传统样本选择方法进行研究的基础上，提出并开发了一种扩展的FCM聚类算法对训练集进行选取以获取高质量的代表性训练数据，从而降低网络异常检测算法的计算开销。采用ETCM-KNN算法的实验表明，在选取相应于原训练集中10％左右的高质量数据进行训练的情况下，仍然能够保证其检测性能没有明显下降，而计算开销要降低70％左右。\n ●提出并研究了使用特征选择及特征加权方法对网络异常检测技术的性能优化策略。通过研究发现，提取和处理的特征数目过多是导致检测速度下降的主要原因之一。基于这个背景，我们将特征选择方法引入入侵检测领域，首先通过大量的对比实验论证了特征选择方法对于提取有效特征、限制特征维数和提高检测速度方面所起到的有效的和积极的作用。然后，我们将其应用于ETCM-KNN异常检测方法当中，实验表明：特征选择方法在保证高检测率、低误报率的前提下，能够极大地降低由于特征过多而带来的过大计算开销(节约开销80％左右)。最后，本文还通过采用基于SVM(支持向量机)的特征加权方法，对使用特征选择技术提取出来的特征进行进一步的加权，从而合理地赋予每个特征对于检测性能的贡献价值。通过使用ETCM-KNN异常检测方法的实验表明，该特征加权方法能够进一步提高算法的检测率(从99.44％提升到99.78％)和降低误报率(从1.74％降低到0.38％)。\n ●通过具体实例探讨了网络异常检测技术的应用前景。基于本文前面的研究成果，将这套整体化、系统化的网络异常检测方法应用于目前非常流行和备受关注的Web服务器的异常检测和统一威胁管理(UTM)系统中的异常检测引擎。从将本

关键词： microbial source tracking   water   machine learning methods   microbial indicators   faecal pollution  

摘要： This paper reports on a successful application of statistical and inductive learning methods to determine optimal discriminating parameters and develop predictive models for the determination of faecal sources in waters, recently and heavily polluted with wastewaters (microbial source tracking). The data comes from an international study in which various microbial and chemical parameters were determined in heavily polluted waters from diverse geographical areas. A total of 38 variables derived from the microbial and chemical parameters were defined to characterise the available 103 observations. Four methods were evaluated: Euclidean k-nearest-neighbour, linear Bayesian classifier, quadratic Bayesian classifier and a support vector machine. The main aim was the obtention of highly accurate predictive models using the lowest number of variables possible. After a strong feature selection process, the obtained results show that predictive models using only two variables emerge with 100% correct classification. The obtained solutions make use of a linear combination of a discriminating tracer (the enumeration of phages infecting Bacteroides thetaiotaomicron) and a universal non-discriminant faecal indicator. Other models not using the discriminant tracer were developed, though a higher number of variables was needed to achieve a high rate of correct classification. (c) 2007 Elsevier Ltd. All rights reserved.