Various policies are being proposed globally to reduce carbon emissions, focusing on improving building energy performance. In order to improve energy performance, it is necessary to conduct an energy assessment of the building. While real testing provides the most accurate results, it is costly and time-consuming. To deal with this, energy simulations are used, but their accuracy often descrease due to simplified parameter. Therefore, studies have been conducted to improve energy performance simulation by increasing the accuracy of variables through sensitivity analysis, or by making corrections to reduce the error value between simulation and test. This study reviews sensitivity analysis methods and bayesian sampling methods and compares their accuracy, computational cost, and practicality in energy performance simulations.