Accurately predicting building energy usage is of great importance in various efforts on improving building energy efficiency. Data-driven approach and first-principle approach are commonly used in developing models for building energy usage prediction. We employed different data driven methods to predict hourly energy usages in two buildings. One was a synthetic large-size office building from DOE reference building models. The other was an existing commercial building. The obtained hourly energy consumption for each building was divided into training and testing sets. We focused on evaluating the accuracy and model complexity of data-driven models for predicting energy usage of two buildings.