Home About us Contact
|
Home About us Contact | ||
|
|
||
Forecast Models (forecast + models)
Selected AbstractsOptimal sampling frequency for volatility forecast models for the Indian stock marketsJOURNAL OF FORECASTING, Issue 1 2009Malay Bhattacharyya Abstract This paper evaluates the performance of conditional variance models using high-frequency data of the National Stock Index (S&P CNX NIFTY) and attempts to determine the optimal sampling frequency for the best daily volatility forecast. A linear combination of the realized volatilities calculated at two different frequencies is used as benchmark to evaluate the volatility forecasting ability of the conditional variance models (GARCH (1, 1)) at different sampling frequencies. From the analysis, it is found that sampling at 30 minutes gives the best forecast for daily volatility. The forecasting ability of these models is deteriorated, however, by the non-normal property of mean adjusted returns, which is an assumption in conditional variance models. Nevertheless, the optimum frequency remained the same even in the case of different models (EGARCH and PARCH) and different error distribution (generalized error distribution, GED) where the error is reduced to a certain extent by incorporating the asymmetric effect on volatility. Our analysis also suggests that GARCH models with GED innovations or EGRACH and PARCH models would give better estimates of volatility with lower forecast error estimates. Copyright © 2008 John Wiley & Sons, Ltd. [source] Comparing density forecast models,JOURNAL OF FORECASTING, Issue 3 2007Yong Bao Abstract In this paper we discuss how to compare various (possibly misspecified) density forecast models using the Kullback,Leibler information criterion (KLIC) of a candidate density forecast model with respect to the true density. The KLIC differential between a pair of competing models is the (predictive) log-likelihood ratio (LR) between the two models. Even though the true density is unknown, using the LR statistic amounts to comparing models with the KLIC as a loss function and thus enables us to assess which density forecast model can approximate the true density more closely. We also discuss how this KLIC is related to the KLIC based on the probability integral transform (PIT) in the framework of Diebold et al. (1998). While they are asymptotically equivalent, the PIT-based KLIC is best suited for evaluating the adequacy of each density forecast model and the original KLIC is best suited for comparing competing models. In an empirical study with the S&P500 and NASDAQ daily return series, we find strong evidence for rejecting the normal-GARCH benchmark model, in favor of the models that can capture skewness in the conditional distribution and asymmetry and long memory in the conditional variance.,,Copyright © 2007 John Wiley & Sons, Ltd. [source] Models to improve winter minimum surface temperature forecasts, Delhi, IndiaMETEOROLOGICAL APPLICATIONS, Issue 2 2004A. P. Dimri Accurate forecasts of minimum surface temperature during winter help in the prediction of cold-wave conditions over northwest India. Statistical models for forecasting the minimum surface temperature at Delhi during winter (December, January and February) are developed by using the classical method and the perfect prognostic method (PPM), and the results are compared. Surface and upper air data are used for the classical method, whereas for PPM additional reanalysis data from the National Center of Environmental Prediction (NCEP) US are incorporated in the model development. Minimum surface temperature forecast models are developed by using data for the winter period 1985,89. The models are validated using an independent dataset (winter 1994,96). It is seen that by applying PPM, rather than the classical method, the model's forecast accuracy is improved by about 10% (correct to within ± 2 °C). Copyright © 2004 Royal Meteorological Society. [source] AN ADAPTIVE LEARNING FRAMEWORK FOR FORECASTING SEASONAL WATER ALLOCATIONS IN IRRIGATED CATCHMENTSNATURAL RESOURCE MODELING, Issue 3 2010SHAHBAZ KHAN Abstract This paper describes an adaptive learning framework for forecasting end-season water allocations using climate forecasts, historic allocation data, and results of other detailed hydrological models. The adaptive learning framework is based on artificial neural network (ANN) method, which can be trained using past data to predict future water allocations. Using this technique, it was possible to develop forecast models for end-irrigation-season water allocations from allocation data available from 1891 to 2005 based on the allocation level at the start of the irrigation season. The model forecasting skill was further improved by the incorporation of a set of correlating clusters of sea surface temperature (SST) and the Southern oscillation index (SOI) data. A key feature of the model is to include a risk factor for the end-season water allocations based on the start of the season water allocation. The interactive ANN model works in a risk-management context by providing probability of availability of water for allocation for the prediction month using historic data and/or with the incorporation of SST/SOI information from the previous months. All four developed ANN models (historic data only, SST incorporated, SOI incorporated, SST-SOI incorporated) demonstrated ANN capability of forecasting end-of-season water allocation provided sufficient data on historic allocation are available. SOI incorporated ANN model was the most promising forecasting tool that showed good performance during the field testing of the model. [source] Jacobian mapping between vertical coordinate systems in data assimilationTHE QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY, Issue 627 2007Y. J. Rochon Abstract Radiances measured by remote-sensing instruments are now the largest component of the atmospheric observation network. The assimilation of radiances from nadir sounders involves fast radiative transfer (RT) models which project profiles provided by forecast models onto the observation space for direct comparison with the measurements. One of the features typically characterizing fast RT models is the use of a fixed vertical coordinate. If the vertical coordinate of the RT model is not identical to that used by the forecast model, an interpolation of forecast profiles to the RT model coordinate is necessary. In variational data assimilation, the mapping of the Jacobians (derivatives of the RT model output with respect to its inputs) from the RT model coordinate to the forecast model coordinate is also required. This mapping of Jacobians is accomplished through the adjoint of the forecast profile interpolator. As shown, the nearest-neighbour log-linear interpolator commonly used operationally can lead to incorrect mapping of Jacobians and, consequently, to incorrect assimilation. This incorrect mapping occurs as a result of leaving out intermediate levels in the interpolation. This problem has been previously masked in part through the smoothing effect of forecast-error vertical correlations on the analysis increments. To solve this problem, two simple versions of an interpolator relying on piecewise log-linear weighted averaging over the layers are investigated. Both markedly improve Jacobian mappings in the assimilation of observations, with one being slightly favoured over the other. This interpolator is being incorporated into the RTTOV model used by several operational weather forecasting centres. Copyright © 2007 Crown in the right of Canada. Published by John Wiley & Sons, Ltd. [source] Constructing a 7-day ahead forecast model for grass pollen at north London, United KingdomCLINICAL & EXPERIMENTAL ALLERGY, Issue 10 2005M. Smith Summary Background A number of media outlets now issue medium-range (,7 day) weather forecasts on a regular basis. It is therefore logical that aerobiologists should attempt to produce medium-range forecasts for allergenic pollen that cover the same time period as the weather forecasts.Objective the objective of this study is to construct a medium-range (7 day) forecast model for grass pollen at north London.Method the forecast models were produced using regression analysis based on grass pollen and meteorological data from 1990 to 1999 and tested on data from 2000 and 2002. The modelling process was improved by dividing the grass pollen season into three periods; the pre-peak, peak and post-peak periods of grass pollen release. The forecast consisted of five regression models: two simple linear regression models predicting the start and end date of the peak period, and three multiple regression models forecasting daily average grass pollen counts in the pre-peak, peak and post-peak periods.Results overall, the forecast models achieved 62% accuracy in 2000 and 47% in 2002, reflecting the fact that the 2002 grass pollen season was of a higher magnitude than any of the other seasons included in the analysis.Conclusion this study has the potential to make a notable contribution to the field of aerobiology. Winter averages of the North Atlantic Oscillation were used to predict certain characteristics of the grass pollen season, which presents an important advance in aerobiological work. The ability to predict allergenic pollen counts for a period between five and seven days will benefit allergy sufferers. Furthermore, medium-range forecasts for allergenic pollen will be of assistance to the medical profession, including allergists planning treatment and physicians scheduling clinical trials. [source] | ||||||||||