mec Utility Analysis uses Linear Regression to establish meter baseline models. It is a statistical technique in which a straight line is fitted to a set of data points (energy usage values) to define the effect of a one or more independent variables such as billing period, heating load, or production volume.
During the baseline modeling process, the energy analyst tries to find the best linear regression fit for all independent variables by varying the heating and cooling balance point temperatures and observing the resulting linear regression.
Generally, the analyst wants to maximize the overall R2 ("R squared") and maximize the T-statistic for each independent variable.