Monitoring and analysis of energy consumption and performance of heating ventilation air conditioning (HVAC) systems in facilities can give insights about building behaviors for improving facility operations. However, due to the custom sensor infrastructure needed to monitor building performance parameters, additional costs would be incurred by owners to instrument facilities and their systems. Detailed analysis of building performance in highly sensed facilities can give insights about the behavior of similar facilities that have no budget to have such sensing infrastructure. This paper provides the analysis of energy consumption and HVAC performance data acquired in a minute interval in a highly sensed building through the comparison of several data mining algorithms, such weighted least square linear regression, random forest and K-nearest neighbor. The findings include the performance of such algorithms in identifying patterns and give insights about suitability of such algorithms in predicting the energy use and system performance in similar buildings with no sensor infrastructure.