Identification of asymmetric prediction intervals through causal forces

International Journal of Forecasting

Collopy

Yokum

2005

Self Cite

Causal forces are a way of summarizing forecasters' expectations about what will happen to a time series in the future. Contrary to the common assumption for extrapolation, time series are not always subject to consistent forces that point in the same direction. Some are affected by conflicting causal forces; we refer to these as complex times series. It would seem that forecasting these times series would be easier if one could decompose the series to eliminate the effects of the conflicts. Given forecasts subject to high uncertainty, we 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. Forecast accuracy for the components can be assessed by testing how well they can be forecast on early holdout data. When such data are not available, historical variability may be an adequate substitute. We tested decomposition by causal forces on 12 complex annual time series for automobile accidents, airline accidents, personal computer sales, airline revenues, and cigarette production. The length of these series ranged from 16 years for airline revenues to 56 years for highway safety data. We made forecasts for one to ten horizons, obtaining 800 forecasts through successive updating. For nine series in which the conditions were completely or partially met, the forecast error (MdAPE) was reduced by more than half. For three series in which the conditions were not met, decomposition by causal forces had little effect on accuracy.

Section: Hypotheses and Prior Researchmentioning

confidence: 99%

Section: Causal Forces To Represent Domain Knowledgementioning

confidence: 99%

Decomposition by causal forces: a procedure for forecasting complex time series

International Journal of Forecasting

Collopy

Yokum

2005

Self Cite

“…Forecasts of these series are likely to be outside the prediction intervals in the direction of the causal forces (Armstrong and Collopy 2000). RBF can highlight these cases for additional attention when making forecasts and estimating prediction intervals.…”

Section: Implications For Practitionersmentioning

confidence: 99%

Rule-Based Forecasting: Using Judgment in Time-Series Extrapolation

International Series in Operations Research &Amp; Management Science

Adya

Collopy

2001

Self Cite

Rule-Based Forecasting (RBF) is an expert system that uses judgment to develop and apply rules for combining extrapolations. The judgment comes from two sources, forecasting expertise and domain knowledge. Forecasting expertise is based on more than a half century of research. Domain knowledge is obtained in a structured way; one example of domain knowledge is managers= expectations about trends, which we call "causal forces. " Time series are described in terms of 28 conditions, which are used to assign weights to extrapolations. Empirical results on multiple sets of time series show that RBF produces more accurate forecasts than those from traditional extrapolation methods or equal-weights combined extrapolations. RBF is most useful when it is based on good domain knowledge, the domain knowledge is important, the series is well behaved (such that patterns can be identified), there is a strong trend in the data, and the forecast horizon is long. Under ideal conditions, the error for RBF's forecasts were one-third less than those for equal-weights combining. When these conditions are absent, RBF neither improves nor harms forecast accuracy. Some of RBF's rules can be used with traditional extrapolation procedures. In a series of studies, rules based on causal forces improved the selection of forecasting methods, the structuring of time series, and the assessment of prediction intervals. Disciplines Business | Marketing ABSTRACTRule-Based Forecasting (RBF) is an expert system that uses judgment to develop and apply rules for combining extrapolations. The judgment comes from two sources, forecasting expertise and domain knowledge. Forecasting expertise is based on more than a half century of research. Domain knowledge is obtained in a structured way; one example of domain knowledge is managers= expectations about trends, which we call "causal forces." Time series are described in terms of 28 conditions, which are used to assign weights to extrapolations. Empirical results on multiple sets of time series show that RBF produces more accurate forecasts than those from traditional extrapolation methods or equal-weights combined extrapolations. RBF is most useful when it is based on good domain knowledge, the domain knowledge is important, the series is well behaved (such that patterns can be identified), there is a strong trend in the data, and the forecast horizon is long. Under ideal conditions, the error for RBF's forecasts were one-third less than those for equal-weights combining. When these conditions are absent, RBF neither improves nor harms forecast accuracy. Some of RBF's rules can be used with traditional extrapolation procedures. In a series of studies, rules based on causal forces improved the selection of forecasting methods, the structuring of time series, and the assessment of prediction intervals.

“…These predictions of annual Ford automobile sales using Holt's extrapolation were obtained from the M-competition study (Makridakis et al 1982, series number 6). We (Armstrong and Collopy 2000) used successive updating over a validation period up to 1967 to calculate the standard 95 percent prediction intervals (dotted lines) from the average ex ante forecast errors for each time horizon. This provided 28 one-year ahead forecasts, 27 two-ahead, and so forth up to 23 six-ahead forecasts.…”

Section: Assessing Uncertaintymentioning

confidence: 99%

“…To determine whether a series is contrary at a given time, we (Armstrong and Collopy 2000) compared the direction implied by the causal forces with the trend component forecasted by Holt's exponential smoothing. We assumed that the causal forces were constant over the forecast horizon for each series.…”

Section: Exhibit 5 Illustration Of Shift In Prediction Intervalsmentioning

confidence: 99%

Extrapolation for Time-Series and Cross-Sectional Data

International Series in Operations Research &Amp; Management Science

2001

Self Cite

Extrapolation methods are reliable, objective, inexpensive, quick, and easily automated. As a result, they are widely used, especially for inventory and production forecasts, for operational planning for up to two years ahead, and for long-term forecasts in some situations, such as population forecasting. This paper provides principles for selecting and preparing data, making seasonal adjustments, extrapolating, assessing uncertainty, and identifying when to use extrapolation. The principles are based on received wisdom (i.e., experts' commonly held opinions) and on empirical studies. Some of the more important principles are:• In selecting and preparing data, use all relevant data and adjust the data for important events that occurred in the past.• Make seasonal adjustments only when seasonal effects are expected and only if there is good evidence by which to measure them.• In extrapolating, use simple functional forms. Weight the most recent data heavily if there are small measurement errors, stable series, and short forecast horizons. Domain knowledge and forecasting expertise can help to select effective extrapolation procedures. When there is uncertainty, be conservative in forecasting trends. Update extrapolation models as new data are received.• To assess uncertainty, make empirical estimates to establish prediction intervals.• Use pure extrapolation when many forecasts are required, little is known about the situation, the situation is stable, and expert forecasts might be biased. Disciplines Business | Marketing ABSTRACTExtrapolation methods are reliable, objective, inexpensive, quick, and easily automated. As a result, they are widely used, especially for inventory and production forecasts, for operational planning for up to two years ahead, and for long-term forecasts in some situations, such as population forecasting. This paper provides principles for selecting and preparing data, making seasonal adjustments, extrapolating, assessing uncertainty, and identifying when to use extrapolation. The principles are based on received wisdom (i.e., experts' commonly held opinions) and on empirical studies. Some of the more important principles are: