多指标可加模型及在医疗费用预测中的应用

对医疗费用的建模分析与合理预测是医疗保险费用厘定的基础与根本.医疗费用中的高维附加信息在长期预测中具有重要作用.然而,传统的统计建模方法不适用于处理高维纵向数据下的医疗费用.本文提出部分线性多指标可加模型,对具有高维特征的纵向医疗费用数据进行拟合与预测,并且使用两种不同的降维估计方法进行模型估计,并将该模型应用于一组含...     

Non-exponential decay kinetics: correct assessment and description illustrated by slow luminescence of Si nanostructures

The treatment of time-resolved (TR) photoluminescence (PL) decay kinetics is analysed in details and illustrated by experiments on semiconductor quantum dots, namely silicon nanocrystals (Si NCs). We consider the mono-, stretch- and multi-exponential as well as lognormal (LN) and some complex decay models for continuous and discrete distribution of rates (lifetimes). A particular attention is devoted to the thorough analysis of non-exponential decay kinetics. We explicitly show that a LN distribution of emitter sizes may results in LN distribution of decay rates. On the other hand, the distribution of rates cannot be, strictly speaking, Levy stable distribution (that results in the stretched-exponential decay). We introduce theoretical background and derive expressions to calculate the average decay lifetimes for some common decays with practical examples of their applications. Experimental aspects are discussed with special attention devoted to the major problems of the accurate TR PL data treatment, including background uncertainty, pulse duration, system response function etc. Finally, a thorough literature survey of TR PL in Si NCs is given. The methods and definitions outlined in this systematic review are applicable to various other material systems with slow decay like rare-earth and transition metal-doped materials, amorphous semiconductors, type-II heterostructures, singlet oxygen phosphorescence etc.     

一种应用于心磁噪声抑制的选择性平均方法研究

在心磁信号探测中, 抑制环境噪声是提取心磁信号的关键. 为了提高心磁信号的信噪比, 信号平均得到了广泛地使用. 然而, 由于局部干扰噪声的存在, 对整段数据进行平均的传统方法不可避免地会带来心磁信号的失真. 本文通过采取模版匹配的方式, 提出了一种选择性平均方法. 结果显示, 本方法能有效剔除带有低频波动、脉冲毛刺等干扰的数据段, 获取高质量的平均心磁信号. 关键词： 心磁 信号平均 噪声抑制 信噪比     

Stationary and multi-self-similar random fields with stochastic volatility

This paper introduces stationary and multi-self-similar random fields which account for stochastic volatility and have type G marginal law. The stationary random fields are constructed using volatility modulated mixed moving average (MA) fields and their probabilistic properties are discussed. Also, two methods for parameterizing the weight functions in the MA representation are presented: one method is based on Fourier techniques and aims at reproducing a given correlation structure, the other method is based on ideas from stochastic partial differential equations. Moreover, using a generalized Lamperti transform we construct volatility modulated multi-self-similar random fields which have type G distribution.     

Risk-sensitive control of continuous time Markov chains

We study risk-sensitive control of continuous time Markov chains taking values in discrete state space. We study both finite and infinite horizon problems. In the finite horizon problem we characterize the value function via Hamilton Jacobi Bellman equation and obtain an optimal Markov control. We do the same for infinite horizon discounted cost case. In the infinite horizon average cost case we establish the existence of an optimal stationary control under certain Lyapunov condition. We also develop a policy iteration algorithm for finding an optimal control.     

Singular control problems in bounded intervals

We consider optimal control problems for one-dimensional diffusion processes y _x (t) = y ^v _x (t), solutions dy _x (t) = g(y _x (t) dt + σ(y _x (t)(dw) + dv _t with y _x(0) = x& isinv;[a,b], the control processes v _t are increasing, positive, and adapted. Several types of expected cost structures associated with each policy v(.) are adopted, e.g. discounted cost, long term average cost and time average cost. Our work is related to [2,6,12,14,16 and 21], where diffusions are allowed to evolve in the whole space, and to [13] and [20], where diffusions evolve only in bounded regions. We shall present some analytic results about value functions, mainly their characterizations, by simple dynamic programming arguments. Several simple examples are explicitly solved to illustrate the singular behaviour of our problems     

A maximal inequality for stochastic convolution integrals on hilbert spaces and space-time regularity of linear stochastic partial differential equations

We consider optimal control problems for one-dimensional diffusion processes [ILM0001] where the control processes υ_t are increasing, positive, and adapted. Several types of expected cost structures associated with each policy υ(.) are adopted, e.g. discounted cost, long term average cost and time average cost. Our work is related to [2,6,12,14,16 and 21], where diffusions are allowed to evolve in the whole space, and to [13] and [20], where diffusions evolve only in bounded regions. We shall present some analytic results about value functions, mainly their characterizations, by simple dynamic programming arguments. Several simple examples are explicitly solved to illustrate the singular behaviour of our problems.     

On circle graphs with girth at least five

Circle graphs with girth at least five are known to be 2-degenerate [A.A. Ageev, Every circle graph with girth at least 5 is 3-colourable, Discrete Math. 195 (1999) 229-233]. In this paper, we prove that circle graphs with girth at least g≥5 and minimum degree at least two contain a chain of g−4 vertices of degree two, which implies Ageev’s result in the case g=5. We then use this structural property to give an upper bound on the circular chromatic number of circle graphs with girth at least g≥5 as well as a precise estimate of their maximum average degree.     

Analysis of a predator-prey model with modified Leslie-Gower and Holling-type II schemes with stochastic perturbation

In this paper, we consider a predator-prey model with modified Leslie-Gower and Holling-type II schemes with stochastic perturbation. We show there is a unique positive solution to the system with positive initial value, and mainly investigate the long time behavior of the system. Condition for the system to be extinct is given and persistent condition is established. At last, numerical simulations are carried out to support our results.     

Sample Average Approximation Method for Chance Constrained Programming: Theory and Applications

We study sample approximations of chance constrained problems. In particular, we consider the sample average approximation (SAA) approach and discuss the convergence properties of the resulting problem. We discuss how one can use the SAA method to obtain good candidate solutions for chance constrained problems. Numerical experiments are performed to correctly tune the parameters involved in the SAA. In addition, we present a method for constructing statistical lower bounds for the optimal value of the considered problem and discuss how one should tune the underlying parameters. We apply the SAA to two chance constrained problems. The first is a linear portfolio selection problem with returns following a multivariate lognormal distribution. The second is a joint chance constrained version of a simple blending problem. B.K. Pagnoncelli’s research was supported by CAPES and FUNENSEG. S. Ahmed’s research was partly supported by the NSF Award DMI-0133943. A. Shapiro’s research was partly supported by the NSF Award DMI-0619977.