Showing papers in "International Journal of Forecasting in 2016"

TL;DR: The need to invest in additional research, such as reproducible case studies, probabilistic load forecast evaluation and valuation, and a consideration of emerging technologies and energy policies in the probabilism load forecasting process are underlined.

TL;DR: This paper introduces the GEFCom2014, a probabilistic energy forecasting competition with four tracks on load, price, wind and solar forecasting, which attracted 581 participants from 61 countries and concludes with 12 predictions for the next decade of energy forecasting.

TL;DR: The mean arctangent absolute percentage error (MAAPE) as discussed by the authors is a new measure of forecast accuracy, which is based on a slope as an angle instead of a slope.

TL;DR: This work presents a technique for the bootstrap aggregation (bagging) of exponential smoothing methods, which results in significant improvements in the forecasts, and demonstrates that it outperforms the originalonential smoothing models consistently.

TL;DR: The methodology of the team Tololo, which ranked first in the load forecasting and price forecasting tracks of the Global Energy Forecasting Competition 2014, is summarized and three methods used during the competition are investigated.

TL;DR: A comprehensive study to model the recency effect using a big data approach and two interesting findings are presented: 1) the naive models are not useful for benchmark purposes in load forecasting at aggregated level due to their lack of accuracy; and 2) slicing the data into 24 pieces to develop one model for each hour is not necessarily better than building one interaction regression model using all 24 hours together.

TL;DR: In this paper, the authors provide a theoretical explanation for the stylized fact that forecast combinations with estimated optimal weights often perform poorly in applications, and there is no guarantee that the optimal forecast combination will be better than the equal weight case, or even improve on the original forecasts.

TL;DR: In this paper, the authors use principal component analysis to automate the process of selecting from among a large set of individual forecasting models that are available for averaging, and show that the resulting Factor Quantile Regression Averaging (FQRA) approach performs very well for price (and load) data from the British power market.

TL;DR: In this article, the authors explore point forecast accuracy measures, explain why 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 non-causal forecasting models on two datasets of daily retail sales.

TL;DR: There are vast numbers of alternative methods for constructing and evaluating value at risk (VaR) forecasts as mentioned in this paper, which are used as a testing ground when fitting alternative models for representing the dynamic evolution of time series of financial returns.

TL;DR: The kernel ridge regression methodology is extended to enable its use for economic time-series forecasting, by including lags of the dependent variable or other individual variables as predictors, as is typically desired in macroeconomic and financial applications.

TL;DR: Stochastic differential equations allow us to capture the time dependence structure of wind speed prediction errors naturally and to derive point and quantile forecasts, predictive distributions, and time-path trajectories, all using one single stochastic differential equation model that is characterized by a few parameters.

TL;DR: This paper used a Markov switching multifractal (MSM) volatility model to forecast crude oil return volatility and found that the model captures stylized facts of multiscaling, long memory, and structural breaks in volatility, after allowing for hundreds of regimes in the volatility.

TL;DR: A voted ensemble of a quantile regression forest model and a stacked random forest–gradient boosting decision tree model is used to predict the probability distribution of solar and wind power generation in connection with the Global Energy Forecasting Competition 2014.

TL;DR: The probabilistic wind power forecasting method used to win the wind track of the Global Energy Forecasting Competition 2014 was consistent throughout the competition, meaning that it can be utilized for similar day-ahead wind forecasting tasks with minimal modeling effort.

TL;DR: In this article, the Mean Absolute Scaled Error (MASE) is proposed as the standard when comparing forecast accuracies, and the MASE fits nicely within the standard statistical procedures initiated by Diebold and Mariano (1995) for testing equal forecast accuracy.

TL;DR: A novel kernel density estimator based on a logarithmic transformation and a boundary kernel is used to construct wind power predictive density based on the k closest historical examples.

TL;DR: The aim of this work is to produce probabilistic forecasts of solar power for the Global Energy Forecasting Competition 2014 using data from the ECMWF numerical weather prediction model.

TL;DR: A probabilistic approach to anomaly detection, specifically in natural gas time series data, which detects anomalies using a linear regression model with weather inputs, after which the anomalies are tested for false positives and classified using a Bayesian classifier.

TL;DR: An integrated solution for probabilistic load forecasting that consists of three components: pre-processing, forecasting, and post-processing and discusses several other variations that were implemented during the GEFCom2014 competition.

TL;DR: In this article, the in-and out-of-sample predictability of US recessions at horizons of three months to two years ahead for a large number of previously proposed leading indicator variables, using the Treasury term spread as a benchmark.

TL;DR: A generic framework for probabilistic energy forecasting is proposed that uses a multiple quantile regression approach to predict a full distribution over possible future energy outcomes, uses the alternating direction method of multipliers to solve the optimization problems resulting from this quantiles regression formulation efficiently, and uses a radial basis function network to capture the non-linear dependencies on the input data.

TL;DR: A methodology for probabilistic load forecasting that is based on lasso (least absolute shrinkage and selection operator) estimation that outperforms two multiple linear regression based benchmarks from among the top eight entries to GEFCom2014-L.

TL;DR: This paper reviews the existing literature on modeling and forecasting call arrivals, and discusses the key issues for the building of good statistical arrival models, and evaluates the forecasting accuracy of selected models in an empirical study with real-life call center data.

TL;DR: This paper proposes a forecasting approach based on a feedforward neural network for probabilistic electricity price forecasting for GEFCom2014 that does not require any special data preprocessing, such as detrending, deseasonality or decomposition of the time series.

TL;DR: This paper provides detailed information on Team Poland’s winning methodology in the electricity price forecasting track of GEFCom2014, and proposes a new hybrid model extending the Quantile Regression Averaging approach of Nowotarski and Weron (2015).

TL;DR: The authors proposed a random projections approach for dealing with the curse of dimensionality issue that afflicts bag-of-words models, which is computationally simple, flexible and fast, and has desirable statistical properties.

TL;DR: For instance, the authors analyzed the predictive benefits associated with the use of principal component analysis, independent component analysis (ICA), and sparse principal component analyses (SPCA) in the context of macroeconomic forecasting.

TL;DR: In this article, the suitability of applying lasso-type penalized regression techniques to macroe-conomic forecasting with high-dimensional datasets was investigated. And the results showed that penalized methods are more robust to mis-specification than factor models, even if the underlying DGP possesses a factor structure.

TL;DR: In this article, a model for generating probabilistic forecasts that combines the kernel density estimation (KDE) and quantile regression techniques, as part of the Probabilistic load forecasting track of the Global Energy Forecasting Competition 2014, is presented.