Search results for "Moving-average model"
showing 9 items of 9 documents
Upport vector machines for nonlinear kernel ARMA system identification.
2006
Nonlinear system identification based on support vector machines (SVM) has been usually addressed by means of the standard SVM regression (SVR), which can be seen as an implicit nonlinear autoregressive and moving average (ARMA) model in some reproducing kernel Hilbert space (RKHS). The proposal of this letter is twofold. First, the explicit consideration of an ARMA model in an RKHS (SVM-ARMA 2k) is proposed. We show that stating the ARMA equations in an RKHS leads to solving the regularized normal equations in that RKHS, in terms of the autocorrelation and cross correlation of the (nonlinearly) transformed input and output discrete time processes. Second, a general class of SVM-based syste…
'Dual' Gravity: Using Spatial Econometrics to Control for Multilateral Resistance
2010
We propose a quantity-based `dual' version of the gravity equation that yields an estimating equation with both cross-sectional interdependence and spatially lagged error terms. Such an equation can be concisely estimated using spatial econometric techniques. We illustrate this methodology by applying it to the Canada-U.S. data set used previously, among others, by Anderson and van Wincoop (2003) and Feenstra (2002, 2004). Our key result is to show that controlling directly for spatial interdependence across trade flows, as suggested by theory, significantly reduces border effects because it captures `multilateral resistance'. Using a spatial autoregressive moving average specification, we …
The predictability of international terrorism: A time‐series analysis
1988
Abstract The study examines the predictability of international terrorism in terms of the existence of trends, seasonality, and periodicity of terrorist events. The data base used was the RAND Corporation's Chronology of International Terrorism. It contains the attributes of every case of international terrorism from 1968 to 1986 (n = 5,589). The authors applied Box‐Jenkins models for a time‐series analysis of the occurrence of terrorist events as well as their victimization rates. The analysis revealed that occurrence of terrorist events is far from being random: There is a clear trend and an almost constant periodicity of one month that can be best described by a first‐order moving averag…
Multi-year drought frequency analysis at multiple sites by operational hydrology - A comparison of methods
2006
Abstract This paper compares two generators of yearly water availabilities from sources located at multiple sites with regard to their ability to reproduce the characteristics of historical critical periods and to provide reliable results in terms of the return period of critical sequences of different length. The two models are a novel multi-site Markov mixture model explicitly accounting for drought occurrences and a multivariate ARMA. In the case of the multisite Markov mixture model parameter estimation is limited to a search in the parameter space guided by the value of parameter λ to show the sensitivity of the model to this parameter. Application to two of the longest time series of …
Tests for time reversibility: a complementarity analysis
2003
Abstract Since time reversibility (TR) is a necessary condition for an independent and identically distributed (iid) sequence, several tests for TR have been suggested to be applied as tests for model misspecification. In this paper, we analyze possible complementarities among two well known TR tests (Ramsey and Rothman's test, and Chen et al.'s test) in two situations: (1) the fitted model is a linear ARMA model when the true data generating process is a nonlinear-in-mean model (either threshold autoregressive or bilinear), and (2) the fitted model is a symmetric GARCH model but the true process belongs to the asymmetric GARCH family (either EGARCH or GJR). The results suggest that there a…
Imputation Strategies for Missing Data in Environmental Time Series for An Unlucky Situation
2005
After a detailed review of the main specific solutions for treatment of missing data in environmental time series, this paper deals with the unlucky situation in which, in an hourly series, missing data immediately follow an absolutely anomalous period, for which we do not have any similar period to use for imputation. A tentative multivariate and multiple imputation is put forward and evaluated; it is based on the possibility, typical of environmental time series, to resort to correlations or physical laws that characterize relationships between air pollutants.
Were the chaotic ELMs in TCV the result of an ARMA process?
2004
The results of a previous paper claiming the demonstration that edge localized mode (ELM) dynamics on TCV are chaotic in a number of cases has recently been called into question, because the statistical test employed was found to also identify linear auto regressive—moving average (ARMA) models as chaotic. The TCV ELM data has therefore been re-examined with an improved method that is able to make this distinction, and the ARMA model is found to be an inappropriate description of the dynamics on TCV. The hypothesis that ELM dynamics are chaotic on TCV in a number of cases is therefore still favoured.
Support Vector Machines Framework for Linear Signal Processing
2005
This paper presents a support vector machines (SVM) framework to deal with linear signal processing (LSP) problems. The approach relies on three basic steps for model building: (1) identifying the suitable base of the Hilbert signal space in the model, (2) using a robust cost function, and (3) minimizing a constrained, regularized functional by means of the method of Lagrange multipliers. Recently, autoregressive moving average (ARMA) system identification and non-parametric spectral analysis have been formulated under this framework. The generalized, yet simple, formulation of SVM LSP problems is particularized here for three different issues: parametric spectral estimation, stability of I…
Spatial moving average risk smoothing
2013
This paper introduces spatial moving average risk smoothing (SMARS) as a new way of carrying out disease mapping. This proposal applies the moving average ideas of time series theory to the spatial domain, making use of a spatial moving average process of unknown order to define dependence on the risk of a disease occurring. Correlation of the risks for different locations will be a function of m values (m being unknown), providing a rich class of correlation functions that may be reproduced by SMARS. Moreover, the distance (in terms of neighborhoods) that should be covered for two units to be found to make the correlation of their risks 0 is a quantity to be fitted by the model. This way, …