基于改进K-means的局部离群点检测方法

离群点检测任务是指检测与正常数据在特征属性上存在显著差异的异常数据。大多数基于聚类的离群点检测方法主要从全局角度对数据集中的离群点进行检测,而对局部离群点的检测性能较弱。基于此,本文通过引入快速搜索和发现密度峰值方法改进K-means聚类算法,提出了一种名为KLOD(local outlier detection based on improved K-means and least-squares methods)的局部离群点检测方法,以实现对局部离群点的精确检测。首先,利用快速搜索和发现密度峰值方法计算数据点的局部密度和相对距离,并将二者相乘得到γ值。其次,将γ值降序排序,利用肘部法则选择γ值最大的k个数据点作为K-means聚类算法的初始聚类中心。然后,通过K-means聚类算法将数据集聚类成k个簇,计算数据点在每个维度上的目标函数值并进行升序排列。接着,确定数据点的每个维度的离散程度并选择适当的拟合函数和拟合点,通过最小二乘法对升序排列的每个簇的每1维目标函数值进行函数拟合并求导,以获取变化率。最后,结合信息熵,将每个数据点的每个维度目标函数值乘以相应的变化率进行加权,得到最终的异常得分,并将异常值得分较高的top-n个数据点视为离群点。通过人工数据集和UCI数据集,对KLOD、LOF和KNN方法在准确度上进行仿真实验对比。结果表明KLOD方法相较于KNN和LOF方法具有更高的准确度。本文提出的KLOD方法能够有效改善K-means聚类算法的聚类效果,并且在局部离群点检测方面具有较好的精度和性能。

On the Equality of Weighted BajratarevićMeans to Quasi-Arithmetic Means

In this paper, we considered the equality problem of weighted Bajraktarević means with weighted quasi-arithmetic means. Using the method of substituting for functions, we first transform the equality problem into solving an equivalent functional equation. We obtain the necessary and sufficient conditions for the equality equation.

An Exploration of Interpersonal Meanings in Advertising Discourse from Functional Grammar Perspective

Interpersonal function is one of the three metafunctions of language. This paper investigates the interpersonal meanings in Chinese cosmetic advertising discourse from mood, modality and appraise the three aspects. Based on an exploratory study of data of six typical Chinese cosmetic advertising discourses, the paper claims that the distribution of mood and modality has close link with the interpersonal meaning in advertising. The Functional Grammar and Appraisal System help to reveal the interpersonal meanings of the discourse concerned.

CONVERGENCE OF A CLASS OF MEANS OF H^p FUNCTIONS(0

In this paper we study the convergence nf a class of means on H^p(G)(0<p<1),the means take the Bochner-Riesz means in[1],the generalized Bochner-Riesz means in[2],and the operators T^(Φ_r)in[3]as special cases.We obtain weak-type estimates for the associated maximal operators and the maximal mean boundedness for the means.

On an Axiomatic about Functional Means

In this paper, we introduce an axiomatic approach about functional means. This includes that of operator means already introduced in the literature.

Application of Multiple Mean Generational Function Method to Typhoon Prediction over the Western North Pacific

[ Objective] The multiple mean generational function （MMGF） method was applied to forecast the annual number of typhoons （TYs） over the Western North Pacific （WNP）. [Method]The method yields a number of predictors by mean generational function based on the rolling 50- year data of TYs frequency and sunspot number, and was repeated to generate forecasts year after year by optimal subset regression. [ Result] The results showed a reasonably high predictive ability dudng period 2000 -2010, with an average root mean square （RMSE） value of 1.92 and a mean absolute error （MAE） value of 1.64. [ Conclusion] Although the MMGF method needs further validation in the practical operation, it already has strong potential for the improvement of skill at forecasting annual frequency of TYs in the WNP.

Some Results on Areally Mean Valent Functions

Let M be the class of areally mean univalent function, f ∈M and In this paper, we estimate the arithmetical mean of coefficients Dn(λ) and the arithmetical mean of successive coefficients tn(λ) =||Dn+1(λ)|-|Dn(λ)||. Our results are sharp. In addition, we also generalize Hayman's theorem on integral mean

Generalized weighted functional proportional mean combining forecasting model and its method of parameter estimation

A new kind of combining forecasting model based on the generalized weighted functional proportional mean is proposed and the parameter estimation method of its weighting coefficients by means of the algorithm of quadratic programming is given. This model has extensive representation. It is a new kind of aggregative method of group forecasting. By taking the suitable combining form of the forecasting models and seeking the optimal parameter, the optimal combining form can be obtained and the forecasting accuracy can be improved. The effectiveness of this model is demonstrated by an example.

APPROXIMATION BY FABER-LAURENT RATIONAL FUNCTIONS IN THE MEAN OF FUNCTIONS OF CLASS LP(T) WITH 1

Let Γ be a regular curve and Lp (Γ), 1<p<+∞. be the class of all complex - valued functions f de-fined on Γ which are such that |f|p is mtegrabie in sense of Lebesgue. In this work, we define the k th p-Fabcrpolynomial F k,p (z),the kth p-Faber principle part F k.p (1/z) for Γ , and defined the nth p-Fcber- Laurent rational function Rn.p (f, z) and p- generalized modulus of continuity Ωp(f, t) of a function f of Lp(Γ) We inves-tigate some properties of Fk,p (z) and Fk.p (1/z). And then we prove a direct theorem characterizing the degree of approximation with respect to Ω (. , t) in the mean of functions of Lp(Γ) by the rational junctions Rn.p (. . z).

APPROXIMATION OF CONTINUOUS FUNCTIONS BY GENERALIZED KARAMATA MEANS

A sufficient condition for the order of approximation of a continuous 2π periodic function with a given majorant for the modulus of continuity by the [F, d_n] means of its Fourier serier to be of Jackson order is obtained. This sufficient condition is shown to be not enough for the order of approximation by partial sums of their Fourier series to be of Jackson order. The error estimate is shown to be the best possible.

基于人工蜂群的三支k-means聚类算法

聚类在数据挖掘技术中起着至关重要的作用。传统的聚类算法都是硬聚类算法,即对象要么属于一个类,要么不属于一个类,在处理不确定数据时,强制划分会带来决策错误。三支k-means聚类算法可以对边界不确定数据进行更加合理的分类,但仍然存在对初始聚类中心敏感的问题。为解决这一问题,将人工蜂群算法与三支k-means聚类算法相结合,提出了一种基于人工蜂群的三支k-means聚类算法。通过定义类内聚集度函数和类间离散度函数来构造蜜源的适应度函数,引导蜂群向高质量的蜜源进行全局搜索。利用蜂群之间不同角色的相互协作与互换,对数据集进行多次迭代聚类,找到最优的蜜源位置,作为初始聚类中心,并在此基础上交替迭代聚类。实验证明,该方法对聚类结果的性能指标有所提高。在UCI数据集上的实验验证了该算法的有效性。

Some Lacunary Sequence Spaces of Invariant Means Defined by Musielak-Orlicz Functions on 2-Norm Space

The purpose of this paper is to introduce and study some sequence spaces which are defined by combining the concepts of sequences of Musielak-Orlicz functions, invariant means and lacunary convergence on 2-norm space. We establish some inclusion relations between these spaces under some conditions.

Pragmatic Functions of Discourse Marker I mean—a case study

The present study examines the pragmatic functions of the discourse marker I mean based on a case study of an American soap opera Desperate Housewives.Four pragmatic functions of I mean are detected and discussed.Among them,two functions are more frequently used including supplementary and explanatory function,reminding and emphasizing function.

Wear State Recognition of Drills Based on K-means Cluster and Radial Basis Function Neural Network

Drill wear not only affects the surface smoothness of the hole, but also influences the life of the drill. Drill wear state recognition is important in the manufacturing process, which consists of two steps： first, decomposing cutting torque components from the original signals by wavelet packet decomposition （WPD）; second, extracting wavelet coefficients of different wear states （i.e., slight, normal, or severe wear） with signal features adapting to Welch spectrum. Finally, monitoring and recognition of the feature vectors of cutting torque signal are performed by using the K-means cluster and radial basis function neural network （RBFNN）. The experiments on different tool wears of the multivariable features reveal that the results of monitoring and recognition are significant and effective.

GAUSS-MEAN VALUE FORMULA FOR TRIHARMONIC FUNCTIONS AND ITS APPLICATIONS IN CLIFFORD ANALYSIS

In this paper, the integral representation for some polyharmonic functions with values in a universal Clifford algebra Cl（Vn,n） is studied and Gauss-mean value formula for triharmonic functions with values in a Clifford algebra Cl（Vn,n） are proved by using Stokes formula and higher order Cauchy-Pompeiu formula. As application some results about growth condition at infinity are obtained.

Sub-Differential Characterizations of Non-Smooth Lower Semi-Continuous Pseudo-Convex Functions on Real Banach Spaces

In this paper, we characterize lower semi-continuous pseudo-convex functions f : X → R ∪ {+ ∞} on convex subset of real Banach spaces K ⊂ X with respect to the pseudo-monotonicity of its Clarke-Rockafellar Sub-differential. We extend the results on the characterizations of non-smooth convex functions f : X → R ∪ {+ ∞} on convex subset of real Banach spaces K ⊂ X with respect to the monotonicity of its sub-differentials to the lower semi-continuous pseudo-convex functions on real Banach spaces.

Mean-square Almost Periodic Random Functions

In this paper, we present a basic theory of mean-square almost periodicity, apply the theory in random differential equation, and obtain mean-square almost periodic solution of some types stochastic differential equation.

设计驱动的产品颠覆性创新方法研究

通过设计驱动创新与颠覆性创新的理论研究,提出了包含意义发掘、功能系统调整、产品语言表达3个阶段的设计驱动产品颠覆性创新过程模型,以完善设计领域对于颠覆性创新方法的研究。首先基于SET洞察市场趋势设想新意义,由内而外接受批评以发掘可行性意义;然后对目标用户进行研究,运用AD将功能需求转化为设计参数,并与源产品技术对比进而进行技术调整,获得颠覆性创新技术解;接着将新意义特征及提升、添加技术所对应的部件确定为意义诠释区域,并采用类比推理选取样本;最后通过相似性分析确定设计原型并提取其特征要素,利用可拓变换优化产品语言表达,获得颠覆性创新产品语言解。以电动轮椅设计为实例验证了该模型的可行性,为其他设计驱动的产品颠覆性创新提供参考。

基于K-means聚类算法的风电光伏光热互补发电机组调度方法

针对光伏光热互补发电机组内部出力不均衡导致的运行不稳、效率低下问题,提出一种基于K-means聚类算法的风电光伏光热互补发电机组调度方法。考虑到光伏光热发电机组具有间歇性、波动性和随机性等特点,采用K-means聚类算法预先对需要调度的数据归类分析,建立光能和风能可能出现的四种组合情况的目标函数,求解函数值,将该值作为下一步调度约束的初始条件值。调度方法结合了功率平衡、蓄能平衡、光伏光热上爬坡及下爬坡事件,计算实时出力值及最佳调度出力值,求解二者差值实现高效调度。试验结果证明,所提方法有效完成了发电机组的电力负荷及功率调度,运行波动和低效问题均得到明显改善,对电站的稳定运行起到了重要作用。

现代汉语名词重叠研究综述

文章主要从普通话、方言及与其他语言的对比这几个角度对现代汉语名词重叠研究进行梳理归纳。通过重点关注研究中主要讨论的重叠式的结构形式、语法意义和语法功能,展现现代汉语名词重叠的研究概况,发现不足并对将来的进一步研究做出展望。