In this paper we will be concerned with the optimisation of non-linear unconstrained problems where in particular the number of variables is large and also the second derivative of the function is sparse. We also present a simple data structure in ADA, that we used in automatic differentiation to obtain the first and second order derivative information of the function. Finally we present some numerical results obtained for solving large dimension problems using the Truncated Newton method.
L. C.W Dixon,
"Experience Using The Truncated Newton Method For Large Scale Optimisation", Proc. SPIE 1058, High Speed Computing II, (17 May 1989); doi: 10.1117/12.951671; https://doi.org/10.1117/12.951671