Ultradistributional boundary values of harmonic functions on the sphere

We present a theory of ultradistributional boundary values for harmonic functions defined on the Euclidean unit ball. We also give a characterization of ultradifferentiable functions and ultradistributions on the sphere in terms of their spherical harmonic expansions. To this end, we obtain explicit estimates for partial derivatives of spherical harmonics, which are of independent interest and refine earlier estimates by Calderón and Zygmund. We apply our results to characterize the support of ultradistributions on the sphere via Abel summability of their spherical harmonic expansions.     

On the support of solutions to stochastic differential equations with path-dependent coefficients

Given a stochastic differential equation with path-dependent coefficients driven by a multidimensional Wiener process, we show that the support of the law of the solution is given by the image of the Cameron–Martin space under the flow of mild solutions to a system of path-dependent ordinary differential equations. Our result extends the Stroock–Varadhan support theorem for diffusion processes to the case of SDEs with path-dependent coefficients. The proof is based on functional Itô calculus.     

Rough paths analysis of general Banach space-valued Wiener processes

In this article, we carry out a rough paths analysis for Banach space-valued Wiener processes. We show that most of the features of the classical Wiener process pertain to its rough path analog. To be more precise, the enhanced process has the same scaling properties and it satisfies a Fernique type theorem, a support theorem and a large deviation principle in the same Hölder topologies as the classical Wiener process does. Moreover, the canonical rough paths of finite dimensional approximating Wiener processes converge to the enhanced Wiener process. Finally, a new criterion for the existence of the enhanced Wiener process is provided which is based on compact embeddings. This criterion is particularly handy when analyzing Kunita flows by means of rough paths analysis which is the topic of a forthcoming article.     

Support vector machines for default prediction of SMEs based on technology credit

In Korea, many forms of credit guarantees have been issued to fund small and medium enterprises (SMEs) with a high degree of growth potential in technology. However, a high default rate among funded SMEs has been reported. In order to effectively manage such governmental funds, it is important to develop an accurate scoring model for selecting promising SMEs. This paper provides a support vector machines (SVM) model to predict the default of funded SMEs, considering various input variables such as financial ratios, economic indicators, and technology evaluation factors. The results show that the accuracy performance of the SVM model is better than that of back-propagation neural networks (BPNs) and logistic regression. It is expected that the proposed model can be applied to a wide range of technology evaluation and loan or investment decisions for technology-based SMEs.     

鲁棒半监督ν-支持向量分类机

支持向量机在近十年成为机器学习的主要学习技术,而且已经成功应用到有监督学习问题中。Fung和Mangasarian利用支持向量机对于既有已标类别样本又有未知类别样本的训练集进行训练,方法主要是利用少量已标明类别的样本进行训练得到一个分类器的同时对于未标明类别的样本进行分类,使得间隔最大化。此优化问题中假定样本是精确的,而在现实生活中,样本通常带有统计误差。因此,考虑样本带有扰动信息的半监督两类分类问题,给出鲁棒半监督v-支持向量分类算法。该算法的参数v易于选择,而数值试验也表明该算法具有良好的稳定性和较好的分类结果。     

线性ε-支持向量回归机的数据扰动分析

支持向量回归机是解决回归问题的一个重要方法.在实际问题中由于测量及计算误差的存在,我们得到的数据往往只是真值的某种近似,带有一定的舍入误差,因此有必要研究支持向量回归机的数据扰动问题.考虑到线性回归问题在实际生活中有广泛的应用价值,把线性ε-支持向量回归机作为研究对象.由于最终关心的是它的原始问题的解,所以我们研究给定的训练集中输入数据发生微小地扰动后,原始问题的解的变化情况.在一定的条件下给出了解对扰动数据偏导数的表达式,建立了线性ε-支持向量回归机的原始问题的灵敏度分析定理.文中还进一步分析了建立该灵敏度分析定理所需要的条件,给出了条件减弱后的结果.文章最后还通过一些简单的数值试验验证了定理的准确性.     

关于Hamilton-Jacobi方程解的正则性和全局结构的注记——纪念李国平院士吴新谋教授诞辰100周年

该文考虑高维Hamilton-Jacobi方程的柯西问题. 作者证明了从任一初始点出发的特征线永不碰到奇异点集合的充分必要条件是初始函数在该点取到最小值.在此基础上,证明了奇异点集合的道路连通分支和初始函数不取最小值的点集合的道路连通分支之间存在一一对应, 而且解的梯度的间断一旦产生就不会消失. 特别指出, 该文的结果不需要“初始函数的梯度在无穷远趋近于零”这一限制条件, 而文献[12]中重要的命题2.7和主要结果之一的定理3.3是在这一条件下得到的.     

贴片电阻表面缺陷自动识别方法

贴片电阻生产过程中的缺陷主要依靠人工在显微镜下检测,速度慢、长期成本高、误检率高.针对贴片电阻单元具有排列整齐、结构简单、图像灰度级少的特点,在贴片电阻图像二值化、边缘提取、直线检测基础上,以相邻电阻单元的相关系数作为电阻缺陷判别依据,提出基于子图投影匹配的快速缺陷检测方法.采用主分量分析法压缩图像数据量,提取缺陷特征,以基于支持向量机对贴片电阻缺陷进行分类并建立实验系统.缺陷检测及识别实验表明,缺陷检测正确率为92.5oo,算法的快速性和识别准确度满足系统快速高精的要求.     

Cohen–Macaulay Modules in Dimension >s and Results on Local Cohomology

Let (R,𝔪) be a local ring and s ≥ ?1. Using the notion of M-sequence in dimension > s, we introduce Cohen–Macaulay modules in dimension > s. Among other things concerning Cohen–Macaulay modules in dimension > s, some finiteness results of the support and the associated primes of local cohomology modules are investigated.     

Automatic Feature Selection via Weighted Kernels and Regularization

Selecting important features in nonlinear kernel spaces is a difficult challenge in both classification and regression problems. This article proposes to achieve feature selection by optimizing a simple criterion: a feature-regularized loss function. Features within the kernel are weighted, and a lasso penalty is placed on these weights to encourage sparsity. This feature-regularized loss function is minimized by estimating the weights in conjunction with the coefficients of the original classification or regression problem, thereby automatically procuring a subset of important features. The algorithm, KerNel Iterative Feature Extraction (KNIFE), is applicable to a wide variety of kernels and high-dimensional kernel problems. In addition, a modification of KNIFE gives a computationally attractive method for graphically depicting nonlinear relationships between features by estimating their feature weights over a range of regularization parameters. The utility of KNIFE in selecting features through simulations and examples for both kernel regression and support vector machines is demonstrated. Feature path realizations also give graphical representations of important features and the nonlinear relationships among variables. Supplementary materials with computer code and an appendix on convergence analysis are available online.