Computational considerations in functional principal component analysis

Computing estimates in functional principal component analysis (FPCA) from discrete data is usually based on the approximation of sample curves in terms of a basis (splines, wavelets, trigonometric functions, etc.) and a geometrical structure in the data space (L ² spaces, Sobolev spaces, etc.). Until now, the computational efforts have been focused in developing ad hoc algorithms to approximate those estimates by previously selecting an efficient approximating technique and a convenient geometrical structure. The main goal of this paper consists of establishing a procedure to formulate the algorithm for computing estimates of FPCA under general settings. The resulting algorithm is based on the classic multivariate PCA of a certain random vector and can thus be implemented in the majority of statistical packages. In fact, it is derived from the analysis of the effects of modifying the norm in the space of coordinates. Finally, an application on real data will be developed to illustrate the so derived theoretic results. This research has been supported by Project MTM2004-5992 from Dirección General de Investigación, Ministerio de Ciencia y Tecnología.     

再生核希尔伯特空间连续线性泛函的范数及其应用北大核心

考虑了再生核希尔伯特空间连续线性泛函范数的表示,得到了用其范数平方等于该线性泛函连续两次作于再生核的简明表示.对于常见的Sobolev-Hibert空间而言,其再生核则可用截幂函数来表示,从而得到Sobolev-Hibert空间上连续线性泛函范数的简洁表示,以新视角解释和简化了文献中的现有结果.     

关于无穷维算子的有限维最佳逼近

A method of approaching to the infinite-dimensional linear operators by the finite-dimensional operators is discussed. It is shown that,for every infinite-dimensional operator A and every natural number n,there exists an n-dimensional optimal approximation to A. The norm error is found and the necessary and sufficient condition for such n-dimensional optimal approximations to be unique is obtained.     

On the Finite-dimensional Optimal Approximations to the Infinite dimensional Linear Operators

§１．　ＩｎｔｒｏｄｕｃｔｉｏｎＬｅｔＨａｎｄＫｂｅｒｅａｌｏｒｃｏｍｐｌｅｘｉｎｆｉｎｉｔｅｄｉｍｅｎｓｉｏｎａｌＨｉｌｂｅｒｓｐａｃｅ．Ａ（ｂｏｕｎｄｅｄｌｉｎｅａｒ）ｏｐ－ｅｒａｔｏｒＡ：Ｈ→Ｋｉｓｓａｉｄｔｏｂｅｉｎｆｉｎｉｔｅ－ｄｉｍｅｎｓｉｏｎａｌ（ｏｒ,ｆｉｎｉｔｅ－ｄｉｍｅｎｓｉｏｎａｌ）ｉｆｔｈｅｒａｎｇｅｏｆＡｉｓｉｎｆｉｎｉｔｅｄｉｍｅｎｓｉｏｎａｌ（ｏｒ,ｆｉｎｉｔｅｄｉｍｅｎｓｉｏｎａｌ）．Ａｉｓｓａｉｄｔｏｂｅｎ－ｄｉｍｅｎｓｉｏｎａｌｉｆｔｈｅｒａｎｇｅｏｆＡｉｓｎ…     

Sharp Conditions for the CLT of Linear Processes in a Hilbert Space

In this paper we study the behavior of sums of a linear process associated to a strictly stationary sequence with values in a real separable Hilbert space and are linear operators from H to H. One of the results is that satisfies the CLT provided are i.i.d. centered having finite second moments and . We shall provide an example which shows that the condition on the operators is essentially sharp. Extensions of this result are given for sequences of weak dependent random variables under minimal conditions.     

关于赋范线性空间中的三类泛函方程

在赋范线性空间中考察下列几类泛函方程( ) f(x) g(y) =h(x+y) ,　 ( ) f(x+y) =f(x) f(y) ,　 ( ) f(x+y) =f(x) +f(y) +ag(x) g(y)的性质与解以及彼此之间的关系 .     

Continuous convergence and functional analysis

The usual setting for Functional Analysis is the category LCS of locally convex topological vector spaces. There are, however, advantages in working in a larger setting, the category CVS of convergence vector spaces—even if one's interest is restricted to LCS. In CVS, one has access to a dual structure, continuous convergence, unavailable in LCS.

We show that theorems such as Grothendieck's completion theorem, Ptak's closed graph and open mapping theorems and the Banach-Steinhaus theorem are transformed from technical results in LCS to transparent and elegant results when examined in CVS with continuous convergence. In the theory of distributions, important bilinear mappings such as evaluations, multiplication and convolution, which are separately continuous when viewed in LCS, become jointly continuous in CVS.     

端单调线性泛函的存在性(英文)

讨论了偏序线性空间的代数对偶空间上的端单调线性泛函的延拓性．进一步,还推导出端单调线性泛函的存在性定理．     

LF拓扑线性空间上Fuzzy线性泛函的连续性

文（４）中给出了Ｘ（Ｌ）上的模糊线性算子的定义。作为其特例，本文证明了ＬＦ拓扑线性空间上模糊线性泛函连续性的几个等价命题和模糊线性泛函的Ｈａｈｎ－Ｂａｎａｃｈ延拓定理，进而给出了ＬＦ拓扑线性空间上存在非零连续模糊线性泛函的一个充要条件。     

Weakly Periodic Sequences of Bounded Linear Transformations: a Spectral Characterization

Let X and Y be two Hilbert spaces, and the space of bounded linear transformations from X into Y. Let {A } be a weakly periodic sequence of period T. Spectral theory of weakly periodic sequences in a Hilbert space is studied by H. L. Hurd and V. Mandrekar (1991). In this work we proceed further to characterize {A _n} by a positive measure and a number T of -valued functions a ₀, . . . ,a _T–1; in the spectral form , where and is an -valued Borel set function on [0, 2) such that      

Invertibility of an Operator Appearing in the Control Theory for Linear Systems

We give sufficient conditions for the existence of a bounded inverse operator for a linear operator appearing in the theory of optimal control of linear systems in Hilbert space and having a matrix representation of the form , where F₃, F₄ are nonnegative self-adjoint operators. The invertibility of the operator under study is used to prove the unique solvability of a certain two-point boundary-value problem that arises from conditions for optimal control.     

Hyperspaces of Normed Linear Spaces with the Attouch–Wets Topology

Let Cld _AW (X) be the hyperspace of nonempty closed subsets of a normed linear space X with the Attouch–Wets topology. It is shown that the space Cld _AW (X) and its various subspaces are AR's. Moreover, if X is an infinite-dimensional Banach space with weight w(X) then Cld _AW (X) is homeomorphic to a Hilbert space with weight 2 ^w(X).     

Functional reproducing kernel Hilbert spaces for non-point-evaluation functional data

Motivated by the need of processing non-point-evaluation functional data, we introduce the notion of functional reproducing kernel Hilbert spaces (FRKHSs). This space admits a unique functional reproducing kernel which reproduces a family of continuous linear functionals on the space. The theory of FRKHSs and the associated functional reproducing kernels are established. A special class of FRKHSs, which we call the perfect FRKHSs, are studied, which reproduce the family of the standard point-evaluation functionals and at the same time another different family of continuous linear (non-point-evaluation) functionals. The perfect FRKHSs are characterized in terms of features, especially for those with respect to integral functionals. In particular, several specific examples of the perfect FRKHSs are presented. We apply the theory of FRKHSs to sampling and regularized learning, where non-point-evaluation functional data are used. Specifically, a general complete reconstruction formula from linear functional values is established in the framework of FRKHSs. The average sampling and the reconstruction of vector-valued functions are considered in specific FRKHSs. We also investigate in the FRKHS setting the regularized learning schemes, which learn a target element from non-point-evaluation functional data. The desired representer theorems of the learning problems are established to demonstrate the key roles played by the FRKHSs and the functional reproducing kernels in machine learning from non-point-evaluation functional data. We finally illustrate that the continuity of linear functionals, used to obtain the non-point-evaluation functional data, on an FRKHS is necessary for the stability of the numerical reconstruction algorithm using the data.     

关于Fuzzy拓扑线性空间的某些结果

本文给出了Ｆｕｚｚｙ拓扑线性空间的若干特征刻划，简化了判断Ｆｕｚｚｙ拓扑线性空间的条件，研究了Ｆｕｚｚｙ拓扑线性空间的层次结构，揭示了Ｆｕｚｚｙ拓扑线性空间与分明拓扑线性空间的内在联系，得到了Ｆｕｚｚｙ拓扑线性空间的“平移不变性”与“局部凸性”都是可截性质。     

赋范线性空间中的Hilbert型积分不等式

基于利用一个积分恒等式的新技巧,建立了赋范线性空间中新的Hilbert型积分不等式.这些新的结果包含了n维欧氏空间中n重积分的Hilbert型积分不等式作为其特殊情形.     

Optimal convergence rates of high order Parzen windows with unbounded sampling

High order Parzen windows are considered with data drawn from unbounded sampling processes. Convergence analysis is established by imposing a moment hypothesis on the unbounded sampling outputs and some decay conditions on the marginal distributions. Our estimate of convergence rate is consistent with the previous result with bounded sampling.     

Functional linear regression with Huber loss

In this paper, we consider the robust regression problem associated with Huber loss in the framework of functional linear model and reproducing kernel Hilbert spaces. We propose an Ivanov regularized empirical risk minimization estimation procedure to approximate the slope function of the linear model in the presence of outliers or heavy-tailed noises. By appropriately tuning the scale parameter of the Huber loss, we establish explicit rates of convergence for our estimates in terms of excess prediction risk under mild assumptions. Our study in the paper justifies the efficiency of Huber regression for functional data from a theoretical viewpoint.     

Application of functional analysis to models of efficient allocation of economic resources

The present paper studies existence and characterization of efficient paths in infinite-horizon economic growth models: the method used is based on techniques of nonlinear functional analysis on Hilbert spaces developed earlier by Chichilnisky. Necessary and sufficient conditions are given for the existence of positive competitive price systems in which the efficient programs maximize present value and intertemporal profit. Approximation of these competitive price systems by strictly positive ones with similar properties is studied. A complete characterization is also given of a class of welfare functions (nonlinear operators defined on consumption paths) for continuity in a weightedl ₂-norm.This research was supported by the National Science Foundation, Grant No. GS-18174. The authors thank K. Arrow, A. Gleason, F. Hahn, A. Majda, S. Marglin, T. Muench, L. Tartar, S. Karamardian, and the referees for helpful comments and suggestions.