Open AccessJournal Article10.3182/20020721-6-ES-1901.00730

Input variable selection for forecasting models

- 01 Jan 2002

- Vol. 35, Iss: 1, pp 463-468

TL;DR: Two approaches to Short Term Load Forecasting generalization are compared using real data from a Spanish utility company and show that input variable selection affects the performance of forecasting models and thus should be treated as a generalization problem.

About: This article is published in IFAC Proceedings Volumes. The article was published on 01 Jan 2002. and is currently open access. The article focuses on the topics: Linear model & Feature selection.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What are the contributions mentioned in the paper "Input variable selection for forecasting models" ?

In this paper two approaches are compared using real data from a Spanish utility company.

2. What future works have the authors mentioned in the paper "Input variable selection for forecasting models" ?

Further research will be aimed at developing neural model testing procedures that would select nets that have the best generalization capabilities.

Fig. 1. Normalized electrical daily load in the working set WS=CS+VS and in the testing set TS.

Table 2. Values of I for the pre-selected variables.

Fig. 5. JT S versus the number of input variables for each model using variables selected by the first method (squares), the second (triangles) and the third method (circles).

Fig. 4. JCS (dark) and JV S (fair gray) for the models derived of the third selection method.

Fig. 3. JCS (dark) and JV S (fair gray) for the models derived of the second selection method.

Table 3. Index for the different variables at each stage of the third selection method.

Citations

Journal Article•10.1016/J.APENERGY.2017.12.051

Predicting electricity consumption for commercial and residential buildings using deep recurrent neural networks

Aowabin Rahman, +2 more

- 15 Feb 2018

- Applied Energy

TL;DR: A recurrent neural network model to make medium-to-long term predictions of electricity consumption profiles in commercial and residential buildings at one-hour resolution and uses the deep NN to perform imputation on an electricity consumption dataset containing segments of missing values is presented.

...read moreread less

675

•Journal Article•10.3390/EN10081168

Short-Term Load Forecasting Using EMD-LSTM Neural Networks with a Xgboost Algorithm for Feature Importance Evaluation

Huiting Zheng, +2 more

- 08 Aug 2017

- Energies

TL;DR: A hybrid algorithm that combines similar days (SD) selection, empirical mode decomposition (EMD), and long short-term memory (LSTM) neural networks to construct a prediction model (i.e., SD-EMD-L STM) for short- term load forecasting is presented.

...read moreread less

627

Journal Article•10.1016/J.ENBUILD.2014.06.052

A prediction model based on neural networks for the energy consumption of a bioclimatic building

Rafael Mena, +3 more

- 01 Oct 2014

- Energy and Buildings

TL;DR: In this paper, a short-term predictive neural network model is proposed to predict the electricity demand for the CIESOL bioclimatic building, located in the southeast of Spain.

...read moreread less

158

•Journal Article•10.3390/EN10010040

Short-Term Forecasting of Electric Loads Using Nonlinear Autoregressive Artificial Neural Networks with Exogenous Vector Inputs

Jaime Buitrago, +1 more

- 01 Jan 2017

- Energies

TL;DR: In this article, a non-linear autoregressive artificial neural network (ANN) with exogenous multi-variable input (NARX) was used for short-term load forecasting.

...read moreread less

112

References

Journal Article•10.1109/72.661122

Neural-network design for small training sets of high dimension

Jen-Lun Yuan, +1 more

- 01 Mar 1998

- IEEE Transactions on Neural Networks

TL;DR: A statistically based methodology for the design of neural networks when the dimension d of the network input is comparable to the size n of the training set, illustrated in detail in the context of short-term forecasting of the demand for electric power from an electric utility.

...read moreread less

Predicción a corto plazo de la demanda de la Energía electríca con SICONEL

Fernando Pavón Pérez, +2 more

- 01 Jan 2000

Proceedings Article•10.23919/ECC.2001.7076354

Neighbor histories for short term electrical load forecasting

Manuel R. Arahal, +1 more

- 01 Sep 2001

TL;DR: The possibility of automatic selection of parameters is investigated to obtain an easily customizable prediction tool and the NH algorithm with the automatically selected parameters is compared against a NN predictor.

...read moreread less

Journal Article•10.1109/59.544636

Neural network based short-term load forecasting using weather compensation

Tommy W. S. Chow, +1 more

- 01 Nov 1996

- IEEE Transactions on Power Systems

TL;DR: In this paper, a nonlinear generalization of Box and Jenkins approach for nonstationary time-series prediction is proposed for electric load forecasting based on neural weather compensation, which can accurately predict the change of actual electric load consumption from the previous day.

...read moreread less

Input variable selection for forecasting models

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What are the contributions mentioned in the paper "Input variable selection for forecasting models" ?

2. What future works have the authors mentioned in the paper "Input variable selection for forecasting models" ?

Figures

Citations

Predicting electricity consumption for commercial and residential buildings using deep recurrent neural networks

Short-Term Load Forecasting Using EMD-LSTM Neural Networks with a Xgboost Algorithm for Feature Importance Evaluation

A prediction model based on neural networks for the energy consumption of a bioclimatic building

Short-Term Forecasting of Electric Loads Using Nonlinear Autoregressive Artificial Neural Networks with Exogenous Vector Inputs

Occupant-centric miscellaneous electric loads prediction in buildings using state-of-the-art deep learning methods

References

Neural-network design for small training sets of high dimension

Predicción a corto plazo de la demanda de la Energía electríca con SICONEL

Neighbor histories for short term electrical load forecasting

Neural network based short-term load forecasting using weather compensation

Related Papers (5)

A general approach based on autocorrelation to determine input variables of neural networks for time series forecasting

Forecasting Macroeconomic Variables using Neural Network Models and Three Automated Model Selection Techniques

Variable Selection Methods for Probabilistic Load Forecasting: Empirical Evidence from Seven States of the United States

Comparing two model selection frameworks for probabilistic load forecasting

Optimal sub-models selection algorithm for combination forecasting model