---

Using decision tree to select forecasting algorithms in distinct electricity consumption context of an office building

The energy optimization of buildings is a crucial challenge that requires advanced solutions for storage and control of expertise. One of the key factors for achieving this goal is the use of forecasting algorithms that can provide accurate and timely predictions of consumption patterns in different contexts. In this paper, it was compared two popular forecasting algorithms: artificial neural networks (ANNs) and k-nearest neighbors (KNNs). These algorithms used consumption data and sensor data in multiple contexts. Decision trees were used to select the best algorithm for each context based on their historical performance metrics. It also investigated how changing the depth parameter of the decision trees affects forecasting accuracy. The results show that the decision trees approach can improve the prediction quality and help to choose the most suitable algorithm for each context.

Publication of the scientific paper "Using decision tree to select forecasting algorithms in distinct electricity consumption context of an office building" in Energy Reports journal (IF: 4.937)). D. Ramos, P. Faria, A. Morais, Z. Vale, “Using decision tree to select forecasting algorithms in distinct electricity consumption context of an office building”, Energy Reports, Volume 8, Supplement 3, 2022, Pages 417-422. DOI: 10.1016/j.egyr.2022.01.046

Open publication