Description of Invention
In this project, a generative adversarial network (GANerAid) is employed to generate data samples from real datasets. The datasets used are collected from MIMOS Berhad. The datasets contain the records of power supply of different combinations from 3 different appliances. GANerAid is used to train and generate synthetic data where data quality is evaluated in terms of root mean square error (RMSE) and visualization. Low scores in RME and the results of visualization indicate that GANerAid could learn effectively from the original datasets to generate accurate synthetic data. GANerAid has shown its usefulness in generating accurate data for machine learning.