Energy Audit System for Households using Machine Learning
A. Nagesh
Dr. A. Nagesh*, Professor, Computer Science & Engineering, Mahatma Gandhi Institute of Technology Hyderabad (Telangana), India.
Manuscript received on May 03, 2021. | Revised Manuscript received on May 08, 2021. | Manuscript published on May 30, 2021. | PP: 33-36 | Volume-10 Issue-7, May 2021 | Retrieval Number: 100.1/ijitee.G88950510721| DOI: 10.35940/ijitee.G8895.0510721
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The growth in population and economics the global demand for energy is increased considerably. The large amount of energy demand comes from houses. Because of this the energy efficiency in houses in considered most important aspect towards the global sustainability. The machine learning algorithms contributed heavily in predicting the amount of energy consumed in household level. In this paper, a energy audit system using machine learning are developed to estimate the amount of energy consumed at household level in order to identify probable areas to plug wastage of energy in household. Each energy audit system is trained using one machine leaning algorithm with previous power consumption history of training data. By converting this data into knowledge, gratification of analysis of energy consumption is attained. The performance of energy audit Linear Regression system is 82%, Decision Tree system is 86% and Random Forest 91% are predicted energy consumption and the performance of learning methods were evaluated based on the heir predictive accuracy, ease of learning and user friendly characteristics. The Random Forest energy audit system is superior when compare to other energy audit system.
Keywords: Energy Prediction, Linear Regression, Machine Learning, Decision Tree, Random Forest.