Demand Analysis of Energy Consumption in a Residential Apartment using Machine Learning

The challenge of keeping pace with the growth of population is synchronizing with the advancement of the technology. As a developing country it is difficult to sustain the balance between the increased inhabitants of Bangladesh and the total energy consumption. Machine learning (ML) can used to forecast the demand of energy consumption including production, organization and conservation of energy for the new buildings with respect to the citizens. This paper introduces a prediction analysis of energy consumption of a residential apartment using different machine learning models including multiple linear regression (MLR), random forest (RF), support vector machine (SVM), and k-nearest neighbors (KNN). While analyzing the models, the energy consumption data of twelve months from an apartment in Chittagong-district situated in Bangladesh was used. The analyses confirm the most effective model used for such energy demand criteria of residential buildings of that location. The outcome of the prediction method reveals that random forest (RF) model can reach to the best accuracy with the highest performance parameters.