PreprintPDF Available
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Many modern statistical models are used for both insight and prediction when applied to data. When models are used for prediction one should optimise parameters through a prediction error loss function. Estimation methods based on multiple steps ahead forecast errors have been shown to lead to more robust and less biased estimates of parameters. However, a plausible explanation of why this is the case is lacking. In this paper, we show that the main benefit of these estimators is in a shrinkage effect, happening in univariate models naturally. However, this can introduce a series of limitations, due to overly aggressive shrinkage. We discuss the predictive likelihoods related to the multistep estimators and demonstrate what their usage implies to time series models. To overcome the limitations of the existing multiple steps estimators, we propose the Geometric Trace Mean Squared Error, demonstrating its advantages. We conduct a simulation experiment showing how the estimators behave with different sample sizes and forecast horizons. Finally, we carry out an empirical evaluation on real data, demonstrating the performance and advantages of the estimators. Given that the underlying process to be modelled is often unknown, we conclude that the GTMSE is a competitive alternative to conventional ones.
Content may be subject to copyright.
A preview of the PDF is not available
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The detection of climate change and its attribution to the corresponding underlying processes is challenging because signals such as trends and shifts are superposed on variability arising from the memory within the climate system. Statistical methods used to characterize change in time series must be flexible enough to distinguish these components. Here we propose an approach tailored to distinguish these different modes of change by fitting a series of models and selecting the most suitable one according to an information criterion. The models involve combinations of a constant mean or a trend superposed to a background of white noise with or without autocorrelation to characterize the memory, and are able to detect multiple changepoints in each model configuration. Through a simulation study on synthetic time series, the approach is shown to be effective in distinguishing abrupt changes from trends and memory by identifying the true number and timing of abrupt changes when they are present. Furthermore, the proposed method is better performing than two commonly used approaches for the detection of abrupt changes in climate time series. Using this approach, the so-called hiatus in recent global mean surface warming fails to be detected as a shift in the rate of temperature rise but is instead consistent with steady increase since the 1960s/1970s. Our method also supports the hypothesis that the Pacific decadal oscillation behaves as a short-memory process rather than forced mean shifts as previously suggested. These examples demonstrate the usefulness of the proposed approach for change detection and for avoiding the most pervasive types of mistake in the detection of climate change.
Article
Full-text available
Supply chain risk management has drawn the attention of practitioners and academics alike. One source of risk is demand uncertainty. Demand forecasting and safety stock levels are employed to address this risk. Most previous work has focused on point demand forecasting, given that the forecast errors satisfy the typical normal i.i.d. assumption. However, the real demand for products is difficult to forecast accurately, which means that---at minimum---the i.i.d. assumption should be questioned. This work analyzes the effects of possible deviations from the i.i.d. assumption and proposes empirical methods based on kernel density estimation (non-parametric) and GARCH(1,1) models (parametric), among others, for computing the safety stock levels. The results suggest that for shorter lead times, the normality deviation is more important, and kernel density estimation is most suitable. By contrast, for longer lead times, GARCH models are more appropriate because the autocorrelation of the variance of the forecast errors is the most important deviation. In fact, even when no autocorrelation is present in the original demand, such autocorrelation can be present as a consequence of the overlapping process used to compute the lead time forecasts and the uncertainties arising in the estimation of the parameters of the forecasting model. Improvements are shown in terms of cycle service level, inventory investment and backorder volume. Simulations and real demand data from a manufacturer are used to illustrate our methodology.
Article
Full-text available
Sales data often only represent a part of the demand for a service product owing to constraints such as capacity or booking limits. Unconstraining methods are concerned with estimating the true demand from such constrained sales data. This paper addresses the frequently encountered situation of observing only a few sales events at the individual product level and proposes variants of small demand forecasting methods to be used for unconstraining. The usual procedure is to aggregate data; however, in that case we lose information on when restrictions were imposed or lifted within a given booking profile. Our proposed methods exploit this information and are able to approximate convex, concave or homogeneous booking curves. Furthermore, they are numerically robust due to our proposed group-based parameter optimization. Empirical results on accuracy and revenue performance based on data from a major car rental company indicate revenue improvements over a best practice benchmark by statistically significant 0.5–1.4% in typical scenarios.
Article
Full-text available
Global surface temperatures continue to rise. In most surface temperature data sets, the years 2014, 2015 and again 2016 set new global heat records since the start of regular measurements. Never before have three record years occurred in a row. We show that this recent streak of record heat does not in itself provide statistical evidence for an acceleration of global warming, nor was it preceded by a 'slowdown period' with a significantly reduced rate of warming. Rather, the data are fully consistent with a steady global warming trend since the 1970s, superimposed with random, stationary, short-term variability. All recent variations in short-term trends are well within what was to be expected, based on the observed warming trend and the observed variability from the 1970s up to the year 2000. We discuss some pitfalls of statistical analysis of global temperatures which have led to incorrect claims of an unexpected or significant warming slowdown.
Article
Massive increases in computing power and new database architectures allow data to be stored and processed at finer and finer granularities, yielding count data time series with lower and lower counts. These series can no longer be dealt with using the approximative methods that are appropriate for continuous probability distributions. In addition, it is not sufficient to calculate point forecasts alone: we need to forecast entire (discrete) predictive distributions, particularly for supply chain forecasting and inventory control, but also for other planning processes. However, tools that are suitable for evaluating the quality of discrete predictive distributions are not commonly used in sales forecasting. We explore classical point forecast accuracy measures, explain why measures such as MAD, MASE and wMAPE are inherently unsuitable for count data, and use the randomized Probability Integral Transform (PIT) and proper scoring rules to compare the performances of multiple causal and noncausal forecasting models on two datasets of daily retail sales.
Article
We extend the argument initiated by Cox (1961) that the exponential smoothing formula can be made more robust for multi-step forecasts if the smoothing parameter is adjusted as a function of the forecast horizon l. The consistency property of the estimator which minimizes the sum of squares of the sample l-step ahead forecast errors makes the robustness result useful in practice. We also generalize the consistency result to some other parsimonious nonstationary models which have been popular in use. The asymptotic distribution of the estimated smoothing parameter adjusted for forecast horizon l leads to the development of diagnostic tools which are based on l-step forecasts.