Currently we minimize bias/absolute errors (annual values); it has been suggested that we consider minimizing differences between modeling results and monthly utility bills (for detailed calibration, for heating/cooling end uses) instead. We would still want to use the bias/absolute errors to determine when we're done searching.
Currently we minimize bias/absolute errors (annual values); it has been suggested that we consider minimizing differences between modeling results and monthly utility bills (for detailed calibration, for heating/cooling end uses) instead. We would still want to use the bias/absolute errors to determine when we're done searching.