Early Prediction on Electrical Energy Consumption in Households by using Machine Learning
Sandy Bhawana Mulia, Ridwan Ridwan, Achmad Ibnu Rosid · 2021
Nowadays Machine Learning is widely used in various fields to perform a smart learning or predictions of a data, one of the methods in Machine Learning is Linear Regression. This research aims to examine the application of Machine Learning with the Linear Regression method programmed with Python Programming to predict electrical energy consumption in the next 30 days for prepaid and postpaid household electricity. Predictions were made based on data collected for 30 days, read through the PZEM-004T V3.0 sensor to measure electrical energy. The comparison experimental results sensor with measurement equipment have an average error of 2% energy reading or with a difference of 0.01 kWh every day for postpaid, and electrical energy consumption with a prediction range of 30 days on prepaid electricity types had an average error value of 1.21% or a difference of 3.68 kWh.