Forecasting time series with outliers
Article Abstract:
The effect of outliers on time-series forecasting accuracy is investigated. Results indicate that the distance of thje outliers from the forecast origin is inversely related to forecast accuracy. To measure the effects of outliers on forecasts, a model is proposed that will facilitate the adjustment for errors. Outlier adjustment improves forecast accuracy if the effect of the outlier is one standard deviation from the innovation series, even if there is a misidentification of the outlier type.
Publication Name: Journal of Forecasting
Subject: Mathematics
ISSN: 0277-6693
Year: 1993
User Contributions:
Comment about this article or add new information about this topic:
A semi-parametric time-series approach in modeling hourly electricity loads
Article Abstract:
The usage of an ARIMA model to derive a backfitting algorithm for electricity load forecasting, based on the non-parametric regression of time series, is described.
Publication Name: Journal of Forecasting
Subject: Mathematics
ISSN: 0277-6693
Year: 2006
User Contributions:
Comment about this article or add new information about this topic:
Optimal prediction with nonstationary ARFIMA model
Article Abstract:
A new methodology for economic forecasting by using long-memory time series with non-stationary ARFIMA model is presented.
Publication Name: Journal of Forecasting
Subject: Mathematics
ISSN: 0277-6693
Year: 2007
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Forecasting the price of crude oil via convenience yield predictions. In search of leading indicators of economic activity in Germany
- Abstracts: Forecasting stock market volatility using (non-linear) Garch models. Do seasonal unit roots matter for forecasting monthly industrial production?
- Abstracts: Communication equilibria with partially verifiable types. Inefficiency of equilibria with incomplete markets. Monetary equilibria in a cash-in-advance economy with incomplete financial markets
- Abstracts: The core can be accessed with a bounded number of blocks. On the continuity of representations of effectivity functions
- Abstracts: Model-based clustering of multiple time series. Binomial autoregressive moving average models with an application to US recessions