关键词： Statistics   Neurosciences   Bayesian statistical decision theory   Calcium imaging  

摘要： This dissertation analyzes how statesmen and soldiers make decisions on war and peace and identifies key linkages between a state's civil-military relations and its international relations. It shows that there is a clear connection between a state's civil-military relations and the making of strategy. This study analyzes how different patterns of civil-military relations affect a state's propensity to use military force, as well as its ability to design effective military strategies to achieve its political objectives. It develops a framework, which allows us to derive hypotheses as to the conditions under which policy-makers would be more likely to make informed decisions on the use of force. The dissertation studies how top decision-makers process information and advice and the political and psychological dynamics that affect the policy-making process. I show that firm civilian control, while necessary, is not a sufficient condition and does not automatically translate into effective national security decision-making or victorious military strategies. Even wealthy and powerful states with firm control over their militaries have paid, at times, enormous cost in blood and treasure due to poorly-conceived diplomatic and military strategies. In the coming years, we are certain to have many occasions when the U.S. uses or considers the use of military force. That is why scholars should continue studying the dynamics of the relations between civilian and military leaders at the pinnacle of government and their effect on the making of military strategy and national security policy.

关键词： 缺省   自然语言处理   句法分析树   机器学习   语料   缺省识别  

摘要： 实现一个基于机器学习的中文缺省项识别系统,对语料库进行预处理,选取多个特征及其组合,通过支持向量模型(SVM)构建的缺省识别模型进行中文缺省识别。研究系统在不同句法分析树上的性能。实验结果证明,该识别系统在标准的句法分析树上F值能达到84.01%,在自动句法树上能达到68.22%。

关键词： machine learning   microscopic review rule   UF-1000i   urinalysis  

摘要： Background: Automated systems have been broadly used in the counting of particles in urine, while manual microscopic analyses are still required for confirming components of urine sediments, especially pathologic casts and other unknown particles. Good review rules can reduce the number of manual urine microscopy examinations safely, thereby increasing productivity. Although several methods have been proposed, establishment of microscopic review rules for flow cytometer remains challenging. Methods: A total of 3014 urine samples from outpatient and inpatient were examined using UF-1000i flow cytometry, Urisys-2400 dipstick and RS 2003 urine sediment workstation, respectively. Based on the results above, three supervised machine learning methods were employed to construct classifiers for screening urine samples. Results: Here, we propose a novel method for construction of microscopic review rules, termed UrineCART, which was based on a classification and regression tree (CART) method. With a cut-off value of 0.0745 for UrineCART, we obtained a sensitivity of 92.0%, a specificity of 81.5% and a total review rate of 32.4% on an independent test set. Comparisons with the existing methods showed that UrineCART gave the acceptable sensitivity and lower total review rate. Conclusions: An algorithm based on machine learning methods for review criteria can be achieved via systematic comparison of UF-1000i flow cytometry and microscopy. Using UrineCART, our microscopic review rate can be reduced to around 30%, while decreasing significant losses in urinalysis.

关键词： Quantitative structure-toxicity relationships   Phenols   Thiophenols   Classification and regression trees   Support vector regression   Photobacterium phosphoreum  

摘要： There is an increasing need for the rapid safety assessment of chemicals by both industries and regulatory agencies throughout the world. In silico techniques are practical alternatives in the environmental hazard assessment. In this background, quantitative structure-toxicity relationship (QSTR) analysis has been performed on toxicity of phenols and thiophenols to Photobacterium phosphoreum. The techniques of classification and regression trees (CART) and least squares support vector regressions (LS-SVR) were applied successfully as variable selection and mapping tools, respectively. Four descriptors selected by the CART technique have been used as inputs of the LS-SVR for prediction of toxicities. The best model explains 91.8% leave-one-out predicted variance and 93.0% external predicted variance. The predictive performance of the CART-LS-SVR model was significantly better than the previous reported models based on CoMFA/CoMSIA and stepwise MLR techniques, suggesting that the present methodology may be useful to predict of toxicity, safety and risk assessment of chemicals. (C) 2012 Elsevier B.V. All rights reserved.

关键词： EEG   K-complex   multi-instance learning (MIL)   sleep disorder  

摘要： Sleep stage identification is the first step in modern sleep disorder diagnostics process. K-complex is an indicator for the sleep stage 2. However, due to the ambiguity of the translation of the medical standards into a computer-based procedure, reliability of automated K-complex detection from the EEG wave is still far from expectation. More specifically, there are some significant barriers to the research of automatic K-complex detection. First, there is no adequate description of K-complex that makes it difficult to develop automatic detection algorithm. Second, human experts only provided the label for whether a whole EEG segment contains K-complex or not, rather than individual labels for each subsegment. These barriers render most pattern recognition algorithms inapplicable in detecting K-complex. In this paper, we attempt to address these two challenges, by designing a new feature extraction method that can transform visual features of the EEG wave with any length into mathematical representation and proposing a hybrid-synergic machine learning method to build a K-complex classifier. The tenfold cross-validation results indicate that both the accuracy and the precision of this proposed model are at least as good as a human expert in K-complex detection.

关键词： 迁移学习   分类   客户流失  

摘要： 数据完备性是基于数据驱动的分析方法的一个重要的前提，不完备的数据意味着很可能会丢失重要的判别信息，从而影响分析结果的准确性。针对现实情况下客户数据特征不同以及数据的不完备性对传统的基于数据驱动的分析方法的不利影响，本文将机器学习领域中迁移分类的方法应用于客户流失预测，通过谱特征排列（spectralfeaturealignment，SFA）实现了跨领域属性的近似统一，并利用直推式支持向量机（transductivesupportvectormachines，TSVM）对客户数据进行分类从而识别忠诚客户和流失客户，使得预测模型的性能显著提高。在最后的实验部分，使用两个不同的数据集进行的迁移分类结果证明了本文提出模型的有效性。

关键词： Grid computing   Wait time prediction   Machine learning  

摘要： To have high performance scheduling mechanisms in grid computing, we need accurate methods for estimating parameters like jobs' wait time and run time. In this paper, we consider wait time prediction problem. Different regression techniques are examined on AuverGrid data set to predict wait time. To improve the quality of prediction, some extra features are proposed. Simulation results show that adding these features reduces prediction error between 13% and 60% in different methods. Results also show that K-nearest neighbor outperforms other regression techniques. We have compared the k-nearest neighbor method in both original and enriched data set with Last-M. K-Nearest neighbor in enriched data set outperforms both Last-M and K-nearest neighbor in original data set in accuracy and perfect prediction percentage.

关键词： Contextual Feature   Text Source   Name Entity Recognition   Machine Learning Classifier   Probabilistic Machine  

摘要： Background: A scientific name for an organism can be associated with almost all biological data. Name identification is an important step in many text mining tasks aiming to extract useful information from biological, biomedical and biodiversity text sources. A scientific name acts as an important metadata element to link biological information. Results: We present NetiNeti (Name Extraction from Textual Information-Name Extraction for Taxonomic Indexing), a machine learning based approach for recognition of scientific names including the discovery of new species names from text that will also handle misspellings, OCR errors and other variations in names. The system generates candidate names using rules for scientific names and applies probabilistic machine learning methods to classify names based on structural features of candidate names and features derived from their contexts. NetiNeti can also disambiguate scientific names from other names using the contextual information. We evaluated NetiNeti on legacy biodiversity texts and biomedical literature (MEDLINE). NetiNeti performs better (precision = 98.9% and recall = 70.5%) compared to a popular dictionary based approach (precision = 97.5% and recall = 54.3%) on a 600-page biodiversity book that was manually marked by an annotator. On a small set of PubMed Central's full text articles annotated with scientific names, the precision and recall values are 98.5% and 96.2% respectively. NetiNeti found more than 190,000 unique binomial and trinomial names in more than 1,880,000 PubMed records when used on the full MEDLINE database. NetiNeti also successfully identifies almost all of the new species names mentioned within web pages. Conclusions: We present NetiNeti, a machine learning based approach for identification and discovery of scientific names. The system implementing the approach can be accessed at http://***.

关键词： Machine learning   Gene expression   Gene regulation   Genetic loci   Gene mapping   Genetic networks   Genetic polymorphism   Machine learning algorithms  

摘要： Expression quantitative trait loci (eQTL) mapping is a widely used technique to uncover regulatory relationships between genes. A range of methodologies have been developed to map links between expression traits and genotypes. The DREAM (Dialogue on Reverse Engineering Assessments and Methods) initiative is a community project to objectively assess the relative performance of different computational approaches for solving specific systems biology problems. The goal of one of the DREAM5 challenges was to reverse-engineer genetic interaction networks from synthetic genetic variation and gene expression data, which simulates the problem of eQTL mapping. In this framework, we proposed an approach whose originality resides in the use of a combination of existing machine learning algorithms (committee). Although it was not the best performer, this method was by far the most precise on average. After the competition, we continued in this direction by evaluating other committees using the DREAM5 data and developed a method that relies on Random Forests and LASSO. It achieved a much higher average precision than the DREAM best performer at the cost of slightly lower average sensitivity.

关键词： Tandem Mass Spectrum   Group Node   Quality Spectrum   Constraint Optimization Problem   Tandem Spectrum  

摘要： Background: In a single proteomic project, tandem mass spectrometers can produce hundreds of millions of tandem mass spectra. However, majority of tandem mass spectra are of poor quality, it wastes time to search them for peptides. Therefore, the quality assessment (before database search) is very useful in the pipeline of protein identification via tandem mass spectra, especially on the reduction of searching time and the decrease of false identifications. Most existing methods for quality assessment are supervised machine learning methods based on a number of features which describe the quality of tandem mass spectra. These methods need the training datasets with knowing the quality of all spectra, which are usually unavailable for the new datasets. Results: This study proposes an unsupervised machine learning method for quality assessment of tandem mass spectra without any training dataset. This proposed method estimates the conditional probabilities of spectra being high quality from the quality assessments based on individual features. The probabilities are estimated through a constraint optimization problem. An efficient algorithm is developed to solve the constraint optimization problem and is proved to be convergent. Experimental results on two datasets illustrate that if we search only tandem spectra with the high quality determined by the proposed method, we can save about 56% and 62% of database searching time while losing only a small amount of high-quality spectra. Conclusions: Results indicate that the proposed method has a good performance for the quality assessment of tandem mass spectra and the way we estimate the conditional probabilities is effective.