Autor: |
Keila Mara Cassiano, Reinaldo Castro Souza, Rafael Morais de Souza, Luiz Albino Teixeira Júnior, Moisés Lima de Menezes |
Rok vydání: |
2014 |
Předmět: |
|
Zdroj: |
Anais do XVII Simpósio de Pesquisa Operacional e Logística da Marinha. |
DOI: |
10.5151/marine-spolm2014-126245 |
Popis: |
Singular Spectrum Analysis (SSA) is a non-parametric technique to decompose a time series into signal and noise. In this article, the Box-Jenkins and Holt-Winters models are tested with and without SSA approach for modeling a time series of monthly residential electricity consumption from a dealership in Rio de Janeiro. Three different methodologies are used in the SSA approach: Analysis of Main Components (ACP), ACP associated with Cluster Analysis and Graphical Analysis of Singular Vectors. MAPE, MAE, RMSE and R2 statistics are used to test the predictive power of the models. The results show a greater predictive power of the model when applied in conjunction with the filtered technique SSA series. |
Databáze: |
OpenAIRE |
Externí odkaz: |
|