基于烟叶致香成分建立烤烟香型分类模型方法研究

为筛选出基于烟叶致香成分数据建立烤烟香型分类的最优模型,以便于较好地对烤烟的香型进行正确分类。首先对142个烤烟烟叶样品中的45个指标采用行业标准进行检测,然后采用逐步回归法筛选出14个烟叶致香成分,依据这14个指标采用判别分析法、Logistic回归、高斯混合模型、分类树、K最邻近法、人工神经网络和支持向量机7种方法进行建模。通过对不同方法建立的模型采用100次随机抽取训练集样本和测试样本计算错误分类率,选择错误分类率较低的模型作为优选模型。结果表明,线性判别法和高斯混和模型建立的2种香型函数能较好地对未知样品的香型进行正确分类,且效果较好(正确率可达90%以上)。研究筛选出的2种优选模型对于烤烟香型分类研究具有一定的应用价值。

基于烟叶化学成分烤烟香型分类模型的建立

基于烟叶化学数据建立烤烟香型分类模型,然后对各模型进行筛选比较选出最优模型。首先对142个烤烟烟叶样品中的9类成分的63个指标采用行业标准进行检测,然后采用逐步回归法筛选出19个烟叶化学成分,依据这19个指标采用线性判别分析法、Logistic回归、高斯混合模型、分类树、K最邻近法、人工神经网络和支持向量机七种方法进行建模。通过对不同方法建立的模型采用100次随机抽取训练集样本和测试样本计算错误分类率,选择错误分类率较低的模型作为优选模型。经比较发现,线性判别法和高斯混合模型建立的两种香型函数能较好地对未知样品的香型进行正确分类,且效果较好。筛选出的两种优选模型对于烤烟香型分类研究具有一定的应用价值。

我国央行冲销成本与冲销干预政策的可持续性研究——基于SVAR模型的实证分析

本文通过SVAR模型检验了我国央行冲销成本与冲销干预政策之间的动态关系。实证分析表明,央行在外汇储备流入加剧时,会通过压制国内利率方式以控制冲销成本,并注意根据冲销成本的高低来把控冲销力度,当冲销成本低时加大冲销力度,在冲销成本较高时减少冲销力度。可见,央行确实考虑了冲销成本对冲销干预政策可持续性的影响。但随着市场化的推进与有利于大规模冲销的制度因素的改革,我国央行控制冲销成本的难度将进一步加大,央行冲销干预政策的可持续性将进一步受到削弱与挑战。

湖北省部分家鸭群体遗传结构探讨

应用16个微卫星标记对湖北省3个家鸭群体共180个个体进行遗传多样性检测。结果表明:湖北省家鸭群体的遗传多样性较为丰富,其中,荆江麻鸭各项指标最高。F-统计量显示,群体间存在明显的遗传差异,有9.7%的遗传变异来自于群体间。Structure程序在预先未标明品种起源的条件下,准确将180个地方鸭个体划分3类,也真实反映了其在遗传上的差异,与各鸭种的选育历史及地理分布基本一致。

Matched Field Localization Based on CS-MUSIC Algorithm

The problem caused by shortness or excessiveness of snapshots and by coherent sources in underwater acoustic positioning is considered.A matched field localization algorithm based on CS-MUSIC(Compressive Sensing Multiple Signal Classification) is proposed based on the sparse mathematical model of the underwater positioning.The signal matrix is calculated through the SVD(Singular Value Decomposition) of the observation matrix.The observation matrix in the sparse mathematical model is replaced by the signal matrix,and a new concise sparse mathematical model is obtained,which means not only the scale of the localization problem but also the noise level is reduced;then the new sparse mathematical model is solved by the CS-MUSIC algorithm which is a combination of CS(Compressive Sensing) method and MUSIC(Multiple Signal Classification) method.The algorithm proposed in this paper can overcome effectively the difficulties caused by correlated sources and shortness of snapshots,and it can also reduce the time complexity and noise level of the localization problem by using the SVD of the observation matrix when the number of snapshots is large,which will be proved in this paper.

Study based on ＂Situational Rationality＂ hypothesis for customer market classification model

The traditional market segmentation was based on ＂transcendental rationality＂ or ＂Situational Rationality＂, studies shows that it had disadvantages. This paper states the ＂Situational＂ integrated rationality hypothesis and then comes up with the market segmenting models and classification algorithm basing on this hypothesis. This algorithm combined the Rough Set theory and Neural Networks in application, which overcome the dilemma that caused complicated network structure and long training time by only using Neural Networks and influenced the classification precision caused by noise disturbance by only using Rough Set methods. Finally, the paper did a comparison experiment between the traditional method and the method we came up, the results shows that the model and algorithm has its advantage on every aspects.

Comparative systems biology between human and animal models based on next-generation sequencing methods

Animal models provide myriad benefits to both experimental and clinical research. Unfortunately, in many situations, they fall short of expected results or provide contradictory results. In part, this can be the result of traditional molecular biological approaches that are relatively inefficient in elucidating underlying molecular mechanism. To improve the efficacy of animal models, a technological breakthrough is required. The growing availability and application of the high-throughput methods make systematic comparisons between human and animal models easier to perform. In the present study, we introduce the concept of the comparative systems biology, which we define as "comparisons of biological systems in different states or species used to achieve an integrated understanding of life forms with all their characteristic complexity of interactions at multiple levels". Furthermore, we discuss the applications of RNA-seq and ChIP-seq technologies to comparative systems biology between human and animal models and assess the potential applications for this approach in the future studies.

Research on Speech Visualization Technology and Its Application in English Listening Teaching

In order to improve the automatic retrieval ability of English vocabulary, for the distribution of semantic attributes in English vocabulary, an automatic classification method of English vocabulary is proposed based on association rules, English vocabulary data storage model is constructed, a two element linguistic feature function is constructed for describing the directionality of English lexical retrieval scheduling, English vocabulary classification decision making model is constructed based on contextual relations of English vocabulary, the features of the association rules of English vocabulary are extracted, the adaptive learning method is used to realize the automatic classification of English vocabulary. The simulation results show that the method of English vocabulary classification has good performance, the classification error rate is low, the retrieval precision is high, and the computational overhead is small.

The art and science of species delimitation

DNA-based approaches to systematics have changed dramatically during the last two decades with the rise of DNA barcoding methods and newer multi-locus methods for species delimitation. During the last half-decade, partly driven by the new sequencing technologies, the focus has shifted to multi-locus sequence data and the identification of species within the frame-work of the multi-species coalescent （MSC）. In this paper, I discuss model-based Bayesian methods for species delimitation that have been developed in recent years using the MSC. Several approximate methods for species delimitation （and their limitations） are also discussed. Explicit species delimitation models have the advantage of clarifying more precisely what is being delimited and what assumptions we are making in doing so. Moreover, the methods can be very powerful when applied to large multi-locus datasets and thus take full advantage of data generated using today＇s technologies [Current Zoology 61 （5）： 846-853,2015].

Network-based naive Bayes model for social network

Naive Bayes(NB) is one of the most popular classification methods. It is particularly useful when the dimension of the predictor is high and data are generated independently. In the meanwhile, social network data are becoming increasingly accessible, due to the fast development of various social network services and websites. By contrast, data generated by a social network are most likely to be dependent. The dependency is mainly determined by their social network relationships. Then, how to extend the classical NB method to social network data becomes a problem of great interest. To this end, we propose here a network-based naive Bayes(NNB) method, which generalizes the classical NB model to social network data. The key advantage of the NNB method is that it takes the network relationships into consideration. The computational efficiency makes the NNB method even feasible in large scale social networks. The statistical properties of the NNB model are theoretically investigated. Simulation studies have been conducted to demonstrate its finite sample performance.A real data example is also analyzed for illustration purpose.

A HYBRID PSO-SA OPTIMIZING APPROACH FOR SVM MODELS IN CLASSIFICATION

Support vector machine （SVM） is a widely used tool in the field of image processing and pattern recognition. However, the parameters selection of SVMs is a dilemma in disease identification and clinical diagnosis. This paper proposed an improved parameter optimization method based on traditional particle swarm optimization （PSO） algorithm by changing the fitness function in the traditional evolution process of SVMs. Then, this PSO method was combined with simulated annealing global searching algorithm to avoid local convergence that traditional PSO algorithms usually run into. And this method has achieved better results which reflected in the receiver-operating characteristic curves in medical images classification and has gained considerable identification accuracy in clinical disease detection.

On the algorithmic construction of classifying spaces and the isomorphism problem for biautomatic groups

We show that the isomorphism problem is solvable in the class of central extensions of word-hyperbolic groups,and that the isomorphism problem for biautomatic groups reduces to that for biautomatic groups with finite centre.We describe an algorithm that,given an arbitrary finite presentation of an automatic group Γ,will construct explicit finite models for the skeleta of K(Γ,1) and hence compute the integral homology and cohomology of Γ.