Package dev.nm.solver.multivariate.unconstrained.c2.steepestdescent
-
Class Summary Class Description FirstOrderMinimizer This implements the steepest descent line search using the first order expansion of the Taylor's series.GaussNewtonMinimizer The Gauss-Newton method is a steepest descent method to minimize a real vector function in the form: /[ f(x) = [f_1(x), f_2(x), ..., f_m(x)]' /] The objective function is /[ F(x) = f' %*% f ]/GaussNewtonMinimizer.MySteepestDescent NewtonRaphsonMinimizer The Newton-Raphson method is a second order steepest descent method that is based on the quadratic approximation of the Taylor series.SteepestDescentMinimizer A steepest descent algorithm finds the minimum by moving along the negative of the steepest gradient direction. -
Enum Summary Enum Description FirstOrderMinimizer.Method the available methods to do line search