Approximation techniques for several stochastic inventory models

This paper considers two methods for approximating several complex stochastic inventory models. The first method involves determining optimal policies for a simpler modified version of the original problem. Examples of this approach, as used by the author, include stocking perishable inventory, the leadtime lost sales problem and an inventory problem with recycling. A second technique, which is appropriate for continuous review models, is based on using a line of reasoning similar to that used by Hadley and Whitin in the development of a heuristic (Q.R) inventory model. Examples of the approach include solutions of some continuous review systems including stocking decaying inventories and determining expected backorders in an inventory system with rationing.     

汉语语音识别中融合发音信息的随机段模型研究

提出了一种基于随机段模型的发音信息集成方法.根据随机段模型的模型特性,建立了阶层式人工神经网络来获取语音段信号属于各类音素的后验概率,并通过一遍解码的方式集成到随机段模型系统中.在“863-test”测试集上进行的汉语连续语音识别实验显示汉语字的相对错误率下降了5.93％.实验结果表明了将发音信息应用到随机段模型的可行性.     

汉语语音识别中融合发音信息的随机段模型研究

提出了一种基于随机段模型的发音信息集成方法。根据随机段模型的模型特性,建立了阶层式人工神经网络来获取语音段信号属于各类音素的后验概率,并通过一遍解码的方式集成到随机段模型系统中。在“863-test”测试集上进行的汉语连续语音识别实验显示汉语字的相对错误率下降了5．93％。实验结果表明了将发音信息应用到随机段模型的可行性。     

语言发音模型研究综述

对各种语言发音模型进行了综述,分别讨论了言语声音模型和言语动作模型。言语声音模型研究语言发音的声学原理,利用声音信号处理技术重构语音信号波形,由于对声源和共鸣之间的关系的认识不同,以及对共鸣的分析方法的不同,产生了3种不同的语言发音模型,第一种是频谱分析模型,第二种是共振峰模型,第三种是生理发音模型。言语动作模型研究发音器官的运动过程,利用图像信号处理技术重构发音器官的发音动作,根据建模方法的不同,言语动作模型可以分为3类：生理机能模型、几何特征模型、统计参数模型。     

Optimization of stochastic simulation models

An algorithm called SAMOPT is developed for optimizing the response function of simulation models that describe systems exhibiting stochastic behavior. Because of the stochastic nature of these simulated systems, the result of each evaluation of response by simulation is only a noisy (i.e., uncertain) observation of the true response. The SAMOPT algorithm uses these noisy responses to find a set of values for decision variables of the system such that the true response is optimized. Principles of the Stochastic Approximation Method have been used in developing this algorithm. The SAMOPT algorithm also allows for the case where the decision variables are subject to a set of linear constraints. Comparison of results between applications of SAMOPT and another well-known method are given for problems and a simulation model.     

面向藏语语音合成的语音基元自动标注方法

在基于隐Markov模型（Hidden Markov Model,HMM）的统计参数藏语语音合成中引入了DAEM（Deterministic Annealing EM）算法,对没有时间标注的藏语训练语音进行自动时间标注。以声母和韵母为合成基元,在声母和韵母的声学模型的训练过程中,利用DAEM算法确定HMM模型的嵌入式重估的最佳参数。训练好声学模型后,再利用强制对齐自动获得声母和韵母的时间标注。实验结果表明,该方法对声母和韵母的时间标注接近手工标注的结果。对合成的藏语语音进行主观评测表明,该方法合成的藏语语音和手工标注声、韵母时间的方法合成的藏语语音的音质接近。因此,利用该方法可以在不需要声、韵母的时间标注的情况下建立合成基元的声学模型。     

Multivariate stochastic fuzzy forecasting models

In this paper, we have presented two new multivariate fuzzy time series forecasting methods. These methods assume m-factors with one main factor of interest. Stochastic fuzzy dependence of order k is assumed to define general methods of multivariate fuzzy time series forecasting and control. These new methods are applied for forecasting total number of car road accidents casualties in Belgium using four secondary factors. Practically, in most of the situations, actuaries are interested in analysis of the patterns of casualties in road accidents. Such type of analysis supports in deciding approximate risk classification and forecasting for each area of a city. This directly affects the underwriting process and adjustment of insurance premium, based on risk intensity for each area. National Institute of Statistics, Belgium provides risk intensity based classification of each city. Thus, this work provides support in deciding the appropriate risk associated with an insured in a particular area.     

Lumping partially symmetrical stochastic models

The performance and dependability evaluation of complex systems by means of dynamic stochastic models (e.g. Markov chains) may be impaired by the combinatorial explosion of their state space. Among the possible methods to cope with this problem, symmetry-based ones can be applied to systems including several similar components. Often however these systems are only partially symmetric: their behavior is in general symmetric except for some local situation when the similar components need to be differentiated.In this paper two methods to efficiently analyze partially symmetrical models are presented in a general setting and the requirements for their efficient implementation are discussed. Some case studies are presented to show the methods’ effectiveness and their applicative interest.     

Closed-form solution of decomposable stochastic models

Markov and semi-Markov processes are increasingly being used in the modeling of complex reconfigurable systems (fault-tolerant computers). The estimation of the reliability (or some measure of performance) of the system reduces to solving the process for its state probabilities. Such a model may exhibit numerous states and complicated transition distributions, contributing to an expensive and numerically delicate solution procedure. Thus, when a system exhibits a decomposition property, either structureally (autonomous subsystems), or behaviorally (component failure versus reconfiguration), it is desirable to exploit this decomposition in the reliability calculation. In interesting cases there can be failure states which arise from non-failure states of the subsystems. We present equation which allow the computation of failure probabilities of the total (combined) model without requiring a complete solution of the combined model. This material is presented within the context of closed-form functional representation of probabilities as utilized in the Symbolic Hierarchical Automated Reliability and Performance Evaluator (SHARPE) tool. The techniques adopted enable one to compute such probability functions for a much wider class of systems at a reduced computational cost. Several examples show how the method is used, especially in enhancing the versatility of the SHARPE tool.     

Transient analysis of stochastic fluid models

We analyze the transient behavior of stochastic fluid flow models in which the input and output rates are controlled by a finite homogeneous Markov process. Such models are used in asynchronous transfer mode (ATM) to evaluate the performance of fast packet switching and in manufacturing systems for the performance of producers and consumers coupled by a buffer. The transient analysis of such models has already been considered in earlier works and solutions have been obtained by the use of Laplace transform. We derive in this paper a new transient solution only based on recurrence relations. We show that this solution is particularly interesting for its numerical properties. The limiting behavior of the solution is also considered. We empirically show that the algorithm for computing the transient solution can be stopped when some stationary behavior is detected.     

Harmonization simulation model for managing several stochastic projects

A highly complicated project management system including several simultaneously realized PERT-COST type network projects, is considered. The projects are of different importance and significance; for each project its corresponding priority index is pre-given. The total budget at the project management disposal to carry out all the projects, is limited. Given for each project its priority value, the problem is to determine optimal budget assignments and optimal due dates of accomplishing each project, to maximize the weighted sum of the projects’ utilities. The problem centers on maximizing the system’s utility by implementing the harmonization model for each single project, which has been outlined in our previous publication [Math. Comput. Simul. (2002)]. Thus, the paper under consideration is an essential extension of that publication.The system’s harmonization model comprises two levels. At the upper level a high-speed lookover search algorithm is implemented, together with a partial harmonization model to determine the projects’ reliability values. At the lower level a linear programming model is imbedded under certain assumptions.     

语音驱动的人脸动画研究现状综述

利用语音来驱动人脸动画,是虚拟现实（Virtual Reality）等领域重要的智能技术,近年来虚拟现实技术的飞速发展更进一步地突出了在沉浸环境下的人机自然交流的迫切需求。语音驱动的人脸动画技术能够创造出自然生动、带有情感的动画,相对于传统预设的人脸动画而言能够更好地辅助人机交互、提升用户体验。为推进该技术的智能化程度和应用,针对语音驱动人脸动画的关键问题：音视频映射,综述了逐帧分析、多帧分析和逐音素分析的映射方法,同时也梳理了多种脸部模型的思想,动画合成、情感融合、人脸动画评价的方法,及可能的研究发展方向。     

噪声鲁棒语音识别研究综述*

针对噪声环境下的语音识别问题,对现有的噪声鲁棒语音识别技术进行讨论,阐述了噪声鲁棒语音识别研究的主要问题,并根据语音识别系统的构成将噪声鲁棒语音识别技术按照信号空间、特征空间和模型空间进行分类总结,分析了各种鲁棒语音识别技术的特点、实现,以及在语音识别中的应用。最后展望了进一步的研究方向。     

Non-intrusive speech quality assessment using several combinations of auditory features

Quality estimation of speech is essential for monitoring and maintenance of the quality of service at different nodes of modern telecommunication networks. It is also required in the selection of codecs in speech communication systems. There is no requirement of the original clean speech signal as a reference in non-intrusive speech quality evaluation, and thus it is of importance in evaluating the quality of speech at any node of the communication network. In this paper, non-intrusive speech quality assessment of narrowband speech is done by Gaussian Mixture Model (GMM) training using several combinations of auditory perception and speech production features, which include principal components of Lyon’s auditory model features, MFCC, LSF and their first and second differences. Results are obtained and compared for several combinations of auditory features for three sets of databases. The results are also compared with ITU-T Recommendation P.563 for non-intrusive speech quality assessment. It is found that many combinations of these feature sets outperform the ITU-T P.563 Recommendation under the test conditions.     

Loss rates for stochastic fluid models

We introduce loss rates, a novel class of performance measures for Markovian stochastic fluid models and discuss their applications potential. We derive analytical expressions for loss rates and describe efficient methods for their evaluation. Further, we study interesting asymptotic properties of loss rates for large size of the buffer, which are crucial for identifying the Quality of Service requirements guaranteed for each user. We illustrate the theory with a numerical example.     

Kalman-Bucy filtering for stochastic Volterra models

In this paper a suboptimization procedure is formulated for Volterra-type equations by using the Kalman-Bucy filtering theory. The Volterra-type systems are, in general, of a time-varying and non-convolution nature, but we include the time-invariant and convolution cases     

On generalised asymmetric stochastic volatility models

Stochastic volatility (SV) models have been considered as a real alternative to time-varying volatility of the ARCH family. Existing asymmetric SV (ASV) models treat volatility asymmetry via the leverage effect hypothesis. Generalised ASV models that take account of both volatility asymmetry and normality violation expressed simultaneously by skewness and excess kurtosis are introduced. The new generalised ASV models are estimated using the Bayesian Markov Chain Monte Carlo approach for parametric and log-volatility estimation. By using simulated and real financial data series, the new models are compared to existing SV models for their statistical properties, and for their estimation performance in within and out-of-sample periods. Results show that there is much to gain from the introduction of the generalised ASV models.