Random Forest Approach for Energy Consumption Behavior Analysis

In today's modern era, with the advent of more sophisticated electrical appliances and their increasing use, energy waste has become one of the most frequently discussed topics. This topic has been featured in the Eleventh Malaysia Plan and has also gained the attention of TNB, which is Malaysia's largest electricity producer. Therefore, energy consumption analysis is needed to identify the behaviour and trends of electricity consumption at particular places and the diversity of their consumers. From this analysis, the energy consumption profile will be developed and can be used to predict daily energy consumption according to respective places and users. Machine learning techniques are commonly used for energy consumption analysis and in particular, Random Forest Classification is the method chosen for this project. To obtain energy data, the Internet of Thing (IoT) technology was adopted to collect energy consumption data, which is then studied for future classification and forecasting of energy consumption. The analysis carried out in this study and their significant findings can bring awareness to consumers and in turn help reduce utility bills.