8. Optimal solution to nonlinear unconstrained problems - steepest descent, Newton's method, FR conjugate gradient
1. Optimization theory
to be continued
2. Steepest descent
Solution steps
Steepest descent procedure diagram
3. Newton iteration method
In order to speed up the convergence, we consider using the second-order Taylor expansion of the objective function f(x) at the iteration point x(k) as the approximation function, and use the minimum point sequence of this quadratic function to approach the minimum point of the objective function.
Newton's iterative method program block diagram