Load Forecasting Using Support Vector Machines: A Study on EUNITE Competition 2001

By Bo-Juen Chen, Ming-Wei Chang, Chih-Jen Lin. 2004

737 citations recorded by the source corpus.

Load forecasting is usually made by constructing models on relative information, such as climate and previous load demand data. In 2001, EUNITE network organized a competition aiming at mid-term load forecasting (predicting daily maximum load of the next 31 days). During the competition we proposed a support vector machine (SVM) model, which was the winning entry, to solve the problem. In this paper, we discuss in detail how SVM, a new learning technique, is successfully applied to load forecasting. In addition, motivated by the competition results and the approaches by other participants, more experiments and deeper analyses are conducted and presented here. Some important conclusions from the results are that temperature (or other types of climate information) might not be useful in such a mid-term load forecasting problem and that the introduction of time-series concept may improve the forecasting.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering, Management Science and Operations Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Electrical and Electronic Engineering

In Management Science and Operations Research

Citations in the corpus

Electric load forecasting methods: Tools for decision making (2009, methodology): “Chen et al. (2004) applied support vector regression for MTLF.”
A Review of Electricity Demand Forecasting in Low and Middle Income Countries: The Demand Determinants and Horizons (2020, methodology): “In the EUNITEComp2001, an SVM model appeared to be the winning entry for medium term forecasts [54].”
Modeling and Forecasting Electricity Loads and Prices: A Statistical Approach (2006, methodology): “Chen et al. (1995) used an adaptive ARMA model for load forecasting, in which the available forecast errors are used to update the model using minimum mean square error to derive error learning coefficients. As reported, the adaptive scheme outperformed conventional ARMA models. Paarmann and Najar’s”
Travel-time prediction with support vector regression (2004, background): “SVM possess great potential and superior performance as is appeared in many previous researches [4][7].”
Energy Forecasting: A Review and Outlook (2020, methodology): “been adopted for energy forecasting for over three decades [24]–[26].”
Time Series Prediction Using Support Vector Machines: A Survey (2009, methodology): “Using MAPE (the same metric used for the EUNITE competition), RMSE, and Max Error metrics, the authors showed that the use of a genetic algorithm to adaptively select the parameters of the SVR outperformed the winners of the EUNITE competition [58, 62].”
Density Forecasting for Long-Term Peak Electricity Demand (2010, methodology): “, 2005); and other methods including machine learning, data mining techniques and grey models (Chen et al., 2004; Fan and Chen, 2006; Morita et al., 1996).”