The standard approach for estimating <a href="search.pl?p=parameters">parameters</a> in a <a href="search.pl?p=regression analysis">regression analysis</a>, based on minimizing the sum of the squared deviations between the actual and fitted values of the criterion (dependent) variable in the <a href="search.pl?p=calibration data">calibration data</a>. (See <a href="search.pl?p=Ordinary Least Squares">Ordinary Least Squares</a>.)<o:p></o:p></P>