Model-based clustering of probability density functions

Complex data such as those where each statistical unit under study is described not by a single observation (or vector variable), but by a unit-specific sample of several or even many observations, are becoming more and more popular. Reducing these sample data by summary statistics, like the average or the median, implies that most inherent information (about variability, skewness or multi-modality) gets lost. Full information is preserved only if each unit is described by a whole distribution. This new kind of data, a.k.a. “distribution-valued data”, require the development of adequate statistical methods. This paper presents a method to group a set of probability density functions (pdfs) into homogeneous clusters, provided that the pdfs have to be estimated nonparametrically from the unit-specific data. Since elements belonging to the same cluster are naturally thought of as samples from the same probability model, the idea is to tackle the clustering problem by defining and estimating a proper mixture model on the space of pdfs. The issue of model building is challenging here because of the infinite-dimensionality and the non-Euclidean geometry of the domain space. By adopting a wavelet-based representation for the elements in the space, the task is accomplished by using mixture models for hyper-spherical data. The proposed solution is illustrated through a simulation experiment and on two real data sets.     

Inverse two-sided Laplace transform for probability density functions

We present a method for the numerical inversion of two-sided Laplace transform of a probability density function. The method assumes the knowledge of the first M derivatives at the origin of the function to be antitransformed. The approximate analytical form is obtained by resorting to maximum entropy principle. Both entropy and L₁-norm convergence are proved. Some numerical examples are illustrated.     

Calculation of probability density functions for nonlinear vibration systems

Technical systems are subjected to a variety of excitations that cannot generally be described in deterministic ways. Random excitations such as road roughness, wind gusts or loads on marine structures are commonly described by stochastic differential equations (SDEs). Given a set of SDEs, the main task is in finding probability density functions (PDFs), which yield statistical information about the system states. Monte-Carlo simulations represent a general way of generating PDFs, however, reliable integration methods can be time-consuming for complex systems. An alternative way of finding PDFs lies in the analysis of the Fokker-Planck equation, a partial differential equation of the PDF. Linear problems under Gaussian excitation can be solved analytically, which is the case only for a small class of nonlinear problems. As a result, there are a number of methods of approximating PDFs for general problems. Most of these are restricted to comparably low dimensions, such as d=4 ("curse of dimensionality"), limiting the relevance to technical applications. This paper presents solutions to problems of dimensions up to d=10, applying a Galerkin-method that expands approximate solutions into orthogonal polynomials. Problems include polynomial nonlinearities in damping and restoring terms, such as classical Duffing-elements, as well as other types of nonlinearities, demonstrated on a typical problem in vehicle dynamics. All results are compared with results from Monte-Carlo simulations or exact solutions, where available. (© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

On adaptive estimation of probability density functions

The paper presents a method of adaptive estimation for a class of probability density functions. This method is a continual analog of some known methods. Bibiligraphy: 10 titles.     

Computing f-divergences and distances of high-dimensional probability density functions

Very often, in the course of uncertainty quantification tasks or data analysis, one has to deal with high-dimensional random variables. Here the interest is mainly to compute characterizations like the entropy, the Kullback–Leibler divergence, more general $f$-divergences, or other such characteristics based on the probability density. The density is often not available directly, and it is a computational challenge to just represent it in a numerically feasible fashion in case the dimension is even moderately large. It is an even stronger numerical challenge to then actually compute said characteristics in the high-dimensional case. In this regard it is proposed to approximate the discretized density in a compressed form, in particular by a low-rank tensor. This can alternatively be obtained from the corresponding probability characteristic function, or more general representations of the underlying random variable. The mentioned characterizations need point-wise functions like the logarithm. This normally rather trivial task becomes computationally difficult when the density is approximated in a compressed resp. low-rank tensor format, as the point values are not directly accessible. The computations become possible by considering the compressed data as an element of an associative, commutative algebra with an inner product, and using matrix algorithms to accomplish the mentioned tasks. The representation as a low-rank element of a high order tensor space allows to reduce the computational complexity and storage cost from exponential in the dimension to almost linear.     

Estimation of the probability density by spline functions

The L_p of a density estimator, constructed with the aid of a spline of degree n, is investigated.Translated from Zapiski Nauchnykh Seminarov Leningradskogo Otdeleniya Matematicheskogo Instituta im. V. A. Steklova AN SSSR, Vol. 166, pp. 32–43, 1988.     

Emptiness formation probability of XX-chain in diffusion process

We study the distribution of emptiness formation probability of XX-model in the diffusion process. There exits a Gaussian decay as well as an exponential decay. The Gaussian decay is caused by the existence of zero point in the Fermi distribution function. The correlation length for each point of scaling factor varies up to the initial condition, monotonically or non-monotonically.     

Approximating probability density functions and their convolutions using orthogonal polynomials

This paper describes and tests methods for piecewise polynomial approximation of probability density functions using orthogonal polynomials. Empirical tests indicate that the procedure described in this paper can provide very accurate estimates of probabilities and means when the probability density function cannot be integrated in closed form. Furthermore, the procedure lends itself to approximating convolutions of probability densities. Such approximations are useful in project management, inventory modeling, and reliability calculations, to name a few applications. In these applications, decision makers desire an approximation method that is robust rather than customized. Also, for these applications the most appropriate criterion for accuracy is the average percent error over the support of the density function as opposed to the conventional average absolute error or average squared error. In this paper, we develop methods for using five well-known orthogonal polynomials for approximating density functions and recommend one of them as giving the best performance overall.     

On the convergence of kernel estimators of probability density functions

Summary The properties of the characteristic function of the fixed-bandwidth kernel estimator of a probability density function are used to derive estimates of the rate of almost sure convergence of such estimators in a family of Hilbert spaces. The convergence of these estimators in a reproducing-kernel Hilbert space is used to prove the uniform convergence of variable-bandwidth estimators. An algorithm employing the fast Fourier transform and heuristic estimates of the optimal bandwidth is presented, and numerical experiments using four density functions are described. This research was supported by the United States Air Force, Air Force Office of Scientific Research, Under Grant Number AFOSR-76-2711.     

Smooth estimation of survival and density functions for a stationary associated process using Poisson weights

Let {X_n,n≥1} be a sequence of stationary non-negative associated random variables with common marginal density f(x). Here we use the empirical survival function as studied in Bagai and Prakasa Rao (1991) and apply the smoothing technique proposed by Gawronski (1980) (see also Chaubey and Sen, 1996) in proposing a smooth estimator of the density function f and that of the corresponding survival function. Some asymptotic properties of the resulting estimators, similar to those obtained in Chaubey and Sen (1996) for the i.i.d. case, are derived. A simulation study has been carried out to compare the new estimator to the kernel estimator of a density function given in Bagai and Prakasa Rao (1996) and the estimator in Buch-Larsen et al. (2005).     

On the Darling-Mandelbrot probability density and the zeros of some incomplete gamma functions

Recently, Mandelbrot has encountered and numerically investigated a probability densityp _d (t) on the nonnegative reals, where, 0D<1. this=" density=" has=" fourier=" transform=">f _d (-is), wheref _d (z)=–Dz ^d (–D, z) and (·.·) is an incomplete gamma function. Previously, Darling had met this density, but had not studied its form. We expressf _d (z) as a confluent hypergeometric function, then locate and approximate its zeros, thereby improving some results of Buchholz. Via properties of Laplace transforms, we approximatep _d (t) asymptotically ast0+ and +, then note some implications asD0+ and 1–.Communicated by Mourad Ismail.     

Moments of the probability density functions of error terms in divisor problems

We derive an explicit formula for the moments of the probability density function of a class of functions. An application of this shows that the density function of the error term in the Pilz divisor problem is asymmetric.

     

Numerical inversion of the Mellin transform on the real line for heavy-tailed probability density functions

A method for numerical inversion on the real line of the Mellin transform, without reduction of the problem to the inversion of Laplace transform is described. Maximum entropy technique is invoked in choosing the analytical form of the approximant function. Entropy-convergence and then L₁-norm convergence is proved. A stability analysis in evaluating entropy and expected values is illustrated. An upper bound of the error in the expected values computation is provided in terms of entropy.     

经济环境下带扩散扰动古典风险过程的破产概率

对于经济环境下带扩散扰动古典风险过程的重要性质进行了讨论,给出了有限时间的破产概率的上界估计.     

Analysis of marginal and conditional density functions for separate inference

Summary This paper discusses, with measure-theoretical rigor, some basic aspects of the theory of separate inference. To analyze densities of marginal and conditional submodels, certain operators are introduced. First a general concept of decomposition of a model is proposed, and the corresponding factorization of densities of the model is established. Next it is shown that the property of smoothness of a family of densities is retained in the operation of conditioning, and therefore it yields the differentiability of the conditional expectation of a real-valued statistic in a certain sense. On the basis of this result, two measures of the effectiveness of a submodel in separate inference are investigated. The Institute of Statistical Mathematics     

Extinction probability for a critical general branching process

A general branching process begins with a single individual born at time t=0. At random ages during its random lifespan L it gives birth to offspring, N(t) being the number born in the age interval [0,t]. Each offspring behaves as a probabilistically independent copy of the initial individual. Let Z(t) be the population at time t, and let N=N(∞). Theorem: If a general branching process is critical, i. e E{N}=1, and if $\text{σ}{}^2\text{=E {N(N?1)}<∞, 0<a≡}\text{∫}\text{0}\text{∞}\text{tdE{N(t)}}$,and as t → ∞ both $\text{t}{}^2\text{(1?}\text{E}\text{{N(t)})→0}$ and $\text{t}{}^2\text{P}\text{[L>t]→0}$, then $\text{tP[Z(t)>0]→}\text{2a}\text{σ}{}^2$ as t→∞.     

Recovering risk-neutral probability density functions from options prices using cubic splines and ensuring nonnegativity

We present a new approach to estimate the risk-neutral probability density function (pdf) of the future prices of an underlying asset from the prices of options written on the asset. The estimation is carried out in the space of cubic spline functions, yielding appropriate smoothness. The resulting optimization problem, used to invert the data and determine the corresponding density function, is a convex quadratic or semidefinite programming problem, depending on the formulation. Both of these problems can be efficiently solved by numerical optimization software.     

Estimation of a probability density function and its derivatives

We investigate statistical estimates of a probability density distribution function and its derivatives. As the starting point of the investigation we take a priori assumptions about the degree of smoothness of the probability density to be estimated. By using these assumptions we can construct estimates of the probability density function itself and its derivatives which are distinguished by the high rate of decrease of the error in the estimate as the sample size increases.Translated from Matematicheskie Zametki, Vol. 12, No. 5, 621–626, November, 1972.     

Complete asymptotic decomposition of the sojourn probability of a diffusion process in thin domains with moving boundaries

We investigate a diffusion process ξ(t) with absorption defined in a thin domainD _ε ={(x,t)∶εG ₁ (t)<x<εG ₂ (t), t≥0}. We obtain the complete decomposition of the sojourn probability of ξ(t) inD _ε with respect to ε→0. Institute of Mathematics, Ukrainian Academy of Sciences, Kiev. Translated from Ukrainskii Matematicheskii Zhurnal, Vol. 51, No. 9, pp. 1155–1164, September, 1999.