Newton correction ($\delta =-DF(x{)}^{-1}F(x)$)

The correction is the vector obtained by solving the Newton system, added to the current point to give $x_{\mathrm{new}}=x+\delta$. The inverse notation specifies the solution; a numerical implementation normally solves a linear system rather than forming a matrix inverse.