Least-squares approach to problems with inexact values of the function and its argument (Q1309050)

The authors consider the problem of reconstructing linear dependence by the least-squares method in the case when both coordinates of the experimental points are known with error. This problem is equivalent to ``least squares fit to a straight line''. A sufficiently simple solution is given in both cases when uncertainties in measurements are unvariable along the line, i.e. they can be described by an operator \(W\), and when the error at \(i\)-th point is characterized by \(W_ i\).