Comparison Between ARIMA Models and K-Nearest Neighbor Method For Forecasting Electricity Consumption in Algeria During the Period (1980-2019)
Journal: Journal of Economic Growth and Entrepreneurship (Vol.3, No. 3)Publication Date: 2020-09-26
Authors : Sahed Abdelkader Kahoui Hacen;
Page : 73-83
Keywords : Forecasting; ARIMA; KNN; Electricity Consumption; Algeria;
- Comparison Between ARIMA Models and K-Nearest Neighbor Method For Forecasting Electricity Consumption in Algeria During the Period (1980-2019)
- Integrating seasonal forecasting using K- nearest neighbor (k-NN) method and CERES-wheat model for management of rainfed wheat cultivation
- DEVELOPMENT OF A CREDIBLE QSAR AND K NEAREST NEIGHBOR MODELS FOR IMIDAZOLYL DERIVATIVES
- Evaluation Accuracy of Nearest Neighbor Sampling Method in Zagross Forests
- THREE-DIMENSIONAL QSAR MODELING BENZIMIDAZOLE ANALOGUES USING THE K-NEAREST NEIGHBOR METHOD
Abstract
The aim of this study is to make a comparison between ARIMA models and the nearest neighbor -K (KNN) method. These models were applied to the annual time series of net electricity consumption in Algeria for the period 1980 to 2019, in a first step a model (0,1,3) (ARIMA) was approved which represents the evolutions of the electricity consumption series, after which the step of applying the method of the nearest neighbor was carried out, where 3 neighbors were identified and the Euclidean distance was calculated between the predicted values and the nearest neighbors, after which the two approaches were compared to show us in the other that the ARIMA model is better and more accurate than the nearest neighbor method, and this is based on the mean square errors MSE and root mean square errors RMSE.
Other Latest Articles
Last modified: 2021-01-26 04:09:19