Newton method (optimization of two variable functions)

The algorithm summarizes Newton's method .

Newton's method uses Hessian and Gradient Information about （ That is, convexity and slope ） To calculate the best point . For most quadratic functions , It only needs one search or 2 The best value can be returned in one iteration , This is even faster than the conjugate gradient method . however , In some cases , For higher order functions or non quadratic functions , The method may diverge or converge to a non minimum fixed point . In order to ensure minimum convergence , A preprocessor is often used . The preprocessor limits the step size , Thus, the calculation amount is increased , But it ensures the smallest solution .

 

Code acquisition