Provided is a dataset preprocessing method for machine learning for energy consumption analysis. According to an embodiment of the present invention, the dataset preprocessing method for machine learning displays measurement data, automatically searches for meaningful intervals and event data requiring analysis from the displayed measurement data, and tags the searched intervals and data. Accordingly, by preprocessing a dataset for machine learning, a data acquisition rate is increased and loss/error data can be easily removed, thereby ultimately increasing the service quality by the analysis function improvement.
展开▼