• Structural Engineering and Mechanics
• /
• v.32 no.1
• /
• pp.71-94
• /
• 2009

In this paper a procedure for Monte Carlo simulation of univariate stationary stochastic processes with the aid of neural networks is presented. Neural networks operate model-free and, thus, circumvent the need of specifying a priori statistical properties of the process, as needed traditionally. This is particularly advantageous when only limited data are available. A neural network can capture the "pattern" of a short observed time series. Afterwards, it can directly generate stochastic process realizations which capture the properties of the underlying data. In the present study a simple feed-forward network with focused time-memory is utilized. The proposed procedure is demonstrated by examples of Monte Carlo simulation, by synthesis of future values of an initially short single process record.

• Communications for Statistical Applications and Methods
• /
• v.26 no.1
• /
• pp.27-34
• /
• 2019

We study the effect of detrending on the coherency between two time series processes. Many economic and financial time series variables include nonstationary components; however, we analyze the two most popular cases of stochastic and deterministic trends. We analyze the asymptotic behavior of coherency under incorrect detrending, which includes the cases of first-differencing the deterministic trend process and, conversely, the time trend removal of the unit root process. A simulation study is performed to investigate the finite sample performance of the sample coherency due to incorrect detrending. Our work is expected to draw attention to the possible distortion of coherency when the series are incorrectly detrended. Further, our results can extend to various specification of trends in aggregate time series variables.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2007.05a
• /
• pp.1437-1440
• /
• 2007

Many hydroclimatic time series are marked by interannual and longer quasi-period features that are associated with narrow band oscillatory climate modes. A time series modeling approach that directly considers such structures is developed and presented. The essence of the approach is to first develop a wavelet decomposition of the time series that retains only the statistically significant wavelet components, and to then model each such component and the residual time series as univariate autoregressive processes. The efficacy of this approach is demonstrated through the simulation of observed and paleo reconstructions of climate indices related to ENSO and AMO, tree ring and rainfall time series. Long ensemble simulations that preserve the spectral attributes of the time series in each ensemble member can be generated. The usual low order statistics are preserved by the proposed model, and its long memory performance is superior to the direction application of an autoregressive model.

• Proceedings of the Korean Statistical Society Conference
• /
• 2002.11a
• /
• pp.265-272
• /
• 2002

An Integer-valued autoregressive integrated (INARI) model is introduced to eliminate stochastic trend and seasonality from time series of count data. This INARI extends the previous integer-valued ARMA model. We show that it is stationary and ergodic to establish asymptotic normality for conditional least squares estimator. Optimal estimating equations are used to reflect categorical and serial correlations arising from panel count data and variations arising from three random processes for obtaining observation into estimation. Under regularity conditions for martingale sequence, we show asymptotic normality for estimators from the estimating equations. Using cancer mortality data provided by the U.S. National Center for Health Statistics (NCHS), we apply our results to estimate the probability of cells classified by 4 causes of death and 6 age groups and to forecast death count of each cell. We also investigate impact of three random processes on estimation.

• Water for future
• /
• v.12 no.2
• /
• pp.63-69
• /
• 1979

The sequences of monthly streamflows constitute a non-statonary time series. The purely stochastic model has been applied to data generation of non-stationary time series. Tow different mothods--single site and multisite generation--have been used on the hydrologic time series. In this study the synthetic generation method by bivariate analysis, studied by Thomas Fiering, one of multi-site models, has been applied to the historical data on monthly streamflows at two sites in Nakdong River, and also for validity of this model the single site Thomas Fiering model applied. Through statistical analysis it has been shown that the performance of bivariate Thomas Fiering model was better than that of the other. By comparison of mean and standard deviaion between the historical and the generated, and cross correlogram interpretation, it has been known that the model used herein has good performance to simultaneously generate the monthly streamflows at two sites in a river hasin.

• Korean Management Science Review
• /
• v.24 no.1
• /
• pp.25-34
• /
• 2007

In this study, we consider stationary waiting times in finite-buffer 3-node single-server queues in series with a Poisson arrival process and with either constant or non-overlapping service times. We assume that each node has a finite buffer except for the first node. The explicit expressions of waiting times in all areas of the stochastic system were driven as functions of finite buffer capacities. These explicit forms show that a system sojourn time does not depend on the finite buffer sizes, and also allow one to compute and compare characteristics of stationary waiting times at all areas under two blocking rules communication and manufacturing blocking. The goal of this study is to apply these results to an optimization problem which determines the smallest buffer capacities satisfying predetermined probabilistic constraints on stationary waiting times at all nodes. Numerical examples are also provided.

• Water for future
• /
• v.11 no.2
• /
• pp.54-61
• /
• 1978

An attemption of synthesizing and forecasting of monthly river flow has been made by employing a linear stochastic difference equation model. As one of the linear stochestic difference equation model, an ARIMA Type is tested to find the suitability of the model to the monthly river flows. On the assumption of the stationary covariacne of differenced monthly river flows the model is identrfield and is evaluated so that the residuale have the minimum variance. Finally a test is performed to finld the residerals beings White noise. Monthly river flows at six stations in Han River Basin are applied for case studies. It was found that the difference operator is a good measure of forecasting the monthly river flow.

• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1201-1210
• /
• 2019

Stock price is characterized as being mutable, non-linear and stochastic. These key characteristics are known to have a direct influence on the stock markets globally. Given that the stock price data often contain both linear and non-linear patterns, no single model can be adequate in modelling and predicting time series data. The autoregressive integrated moving average (ARIMA) model cannot deal with non-linear relationships, however, it provides an accurate and effective way to process autocorrelation and non-stationary data in time series forecasting. On the other hand, the neural network provides an effective prediction of non-linear sequences. As a result, in this study, we used a hybrid ARIMA and neural network model to forecast the monthly closing price of the Shanghai composite index and Shenzhen component index.

• Management Science and Financial Engineering
• /
• v.14 no.1
• /
• pp.23-34
• /
• 2008

In this study, we consider stationary waiting times in a Poisson driven single-server m-node queues in series. We assume that service times at nodes are independent, and are either deterministic or non-overlapped. Each node excluding the first node has a finite waiting line and every node is operated under a FIFO service discipline and a communication blocking policy (blocking before service). By applying (max, +)-algebra to a corresponding stochastic event graph, a special case of timed Petri nets, we derive the explicit expressions for stationary waiting times at all areas, which are functions of finite buffer capacities. These expressions allow us to compute the performance measures of interest such as mean, higher moments, or tail probability of waiting time. Moreover, as applications of these results, we introduce optimization problems which determine either the biggest arrival rate or the smallest buffer capacities satisfying probabilistic constraints on waiting times. These results can be also applied to bounds of waiting times in more general systems. Numerical examples are also provided.

• Journal of the Korean Statistical Society
• /
• v.26 no.3
• /
• pp.407-416
• /
• 1997

Suppose that { $X_{i}$ } is a stationary AR(1) process and { $Y_{j}$ } is an ARX process with { $X_{i}$ } as exogeneous variables. Let $Y_{j}$ $^{*}$ be the stochastic process which is the sum of $Y_{j}$ and a nonstochastic trend. In this paper we consider the problem of estimating the conditional probability that $Y_{{n+1}}$$^{*}$ is bigger than $X_{{n+1}}$, given $X_{1}$, $Y_{1}$$^{*}$,..., $X_{n}$ , $Y_{n}$ $^{*}$. As an estimator for the tolerance probability, an Mann-Whitney statistic based on least squares residuars is suggested. It is shown that the deviations between the estimator and true probability are stochatically bounded with $n^{{-1}$2}/ order. The result may be applied to the stress-strength reliability theory when the stress and strength variables violate the classical iid assumption.umption.n.