Showing papers in "International Journal of Forecasting in 2005"

TL;DR: In this article, forecasting techniques to predict the 24 market-clearing prices of a day-ahead electric energy market were considered, including time series analysis, neural networks and wavelets, and extensive analysis was conducted using data from the PJM Interconnection.

TL;DR: In this article, four forecasting methods, Simple Moving Average (SMA), 13 periods, Single Exponential Smoothing (SES), and a new method (based on Croston's approach) were compared on 3000 real intermittent demand data series from the automotive industry.

TL;DR: In this paper, the authors examined the predictability of stock returns using macroeconomic variables in 12 industrialized countries and employed recently developed out-of-sample tests that have increased power, namely, the McCracken [ Asymptotics for out-oftheoretic tests of Granger Causality, Manuscript, University of Missouri-Columbia (2004) and the West [ Econometrica 64 (1996) 1067] test for equal predictive ability.

TL;DR: In this article, a simple model-selection criterion was proposed to select among forecasts, and the accuracy of the selected combinations is significantly better and less variable than that of selected individual forecasts.

TL;DR: The authors examined the forecast accuracy of linear autoregressive, smooth transition auto-regression (STAR), and neural network (NN) time series models for 47 macroeconomic variables of the G7 economies.

TL;DR: In this paper, the relative out-of-sample predictive ability of different GARCH models, with particular emphasis on the predictive content of the asymmetric component, was examined and compared.

TL;DR: A dynamic neural network model for forecasting time series events that uses a different architecture than traditional models is presented and shows that this approach is more accurate and performs significantly better than the traditional neural network and autoregressive integrated moving average (ARIMA) models.

TL;DR: In this article, the authors compare the effectiveness of forecasts based on published odds and forecasts made using a benchmark statistical model incorporating a large number of quantifiable variables relevant to match outcomes.

TL;DR: In this paper, Bivariate Poisson regression is used to estimate forecasting models for goals scored and conceded, and Ordered Probit Regression (OBR) is used for match results.

TL;DR: A new methodology that has not previously been used to evaluate economic forecasts: multiple comparisons is introduced, which concludes that the accuracy of the various methods does differ significantly, and that some methods are significantly better than others.

TL;DR: In this paper, a short-term forecasting model of monthly West Texas Intermediate crude oil spot prices using readily available OECD industrial petroleum inventory levels is presented, which provides good in-sample and out-of-sample dynamic forecasts for the post-Gulf War time period.

TL;DR: In this article, the authors examined whether data from business tendency surveys are useful for forecasting GDP growth in the short run, using a so-called dynamic factor model (DFM), which is used both as a framework for dimension reduction in forecasting and as a procedure for filtering out unimportant idiosyncratic noise in the underlying survey data.

TL;DR: In this paper, the authors analyzed whether the accuracy of forecasts of aggregate euro area inflation can be improved by aggregating forecasts of subindices of the Harmonized Index of Consumer Prices (HICP) as opposed to forecasting the aggregate HICP directly.

TL;DR: Comparisons of several forecasting methods on a sample of hourly electricity demands, including both large neural networks and conventional regression-based methods, find good performance for the large Neural Network, and some analysis of why forecasting the 24 element vector of daily electricity demands may be particularly conducive to this approach.

TL;DR: In this paper, the authors evaluate the asymptotic and finite-sample properties of direct multi-step estimation for forecasting at several horizons, and show that if a model is mis-specified for a non-stationary DGP, in particular omitting either negative residual serial correlation or regime shifts, DMS can forecast more accurately.

TL;DR: In this article, the p-step ahead predictive mass functions for a family of distributions nested in the integer-valued first-order autoregressive (INAR(1)) class are constructed from convolutions of the unobserved components of the model, with uncertainty associated with both parameter values and model specification fully incorporated.

TL;DR: This article investigated the forecasting performance and confidence of experts and non-experts in the first round of World Cup 2002 and found that experts overestimated their performance and tended to be overconfident, while the opposite tendency was observed for the participants with limited knowledge.

TL;DR: In this paper, the authors applied linear regression as well as nonlinear models to examine the predictive accuracy of the term spread-output growth relationship and found significant nonlinearity with respect to time and past annual growth.

TL;DR: In this article, the forecasting performance of various models for seasonality and nonlinearity for quarterly industrial production series of 18 OECD countries was examined and it was found that linear models with fairly simple descriptions of seasonality outperformed nonlinear models with more elaborate seasonal components.

TL;DR: In this paper, two rating philosophies are distinguished: through the cycle versus point in time ratings and correlations derived from a nonlinear random effects panel model using data from Standard & Poor's.

TL;DR: In this paper, the authors report the results of fitting unobserved components (structural) time series models to data on real income per capita in eight regions of the United States.

TL;DR: The authors found that game theorists' forecasts were less accurate than forecasts from student role players in simulated interactions, and that experienced game theorists were not more accurate than inexperienced role-players in the task.

TL;DR: In this article, the authors construct prediction intervals for autoregressive conditional heteroskedasticity (ARCH) models using the bootstrap and compare their prediction intervals to traditional asymptotic prediction intervals.

TL;DR: Results indicate that people benefit from the use of the decomposition-based decision aid in the task, but, unexpectedly, there was no greater benefit when information load was greatest.

TL;DR: In this paper, the authors show that for a large class of a priori designed filters, an AMB interpretation is always possible, and that proper convolution of AMB filters can produce richer decompositions of the series that fully respect the ARIMA model for the observed series.

TL;DR: In this paper, a multivariate El Himdi-Roy (HR) test is adapted to jointly test the forecasting value of multiple production expectation series, to assess whether part of this joint effect is indeed due to cross-country influences, and to determine which countries' expectation series have the most "clout" in predicting the production levels in other member countries, or have the highest receptivity, in that their production levels are Granger caused by the other countries' expectations.

TL;DR: In this paper, the authors build monthly coincident and leading composite indicators for the euro area growth cycle using frequency domain analysis to achieve additional insight about their relationships at different frequency bands.

TL;DR: This article reported a laboratory study on the reactions of forecasters to different types of loss functions, where subjects were given a cover story that they were the production manager in an organization with an asymmetric loss function.

TL;DR: It is hypothesized that a time series could be effectively decomposed under two conditions: 1) if domain knowledge can be used to structure the problem so that causal forces are consistent for two or more component series, and 2) when it is possible to obtain relatively accurate forecasts for each component.

TL;DR: In this paper, model selection criteria are applied for finding the order of the best autoregressive model fitted to the squared residuals of the linear model, and if the order selected is not zero, this is considered as an indication of nonlinear behavior.