# What is the fuzzy search algorithm faster?

What is the fuzzy search algorithm faster?
September 19th 19 at 13:20
September 19th 19 at 13:22
The algorithm is Monte Carlo the fastest.
tag < sarcasm > - jaren.Rutherford commented on September 19th 19 at 13:40
He can't take a joke - earnestine.Smitham72 commented on September 19th 19 at 13:25
I do not understand, You exactly can you write? - Araceli.Gorczany commented on September 19th 19 at 13:28
Specifically. Toss a coin if it's heads-to return a cookie, if tails -- cats. Users will enjoy.

Seriously, if you want a good answer, put the question adequately. What you're looking for? What are you looking for? What is the metric of similarity? Whether changing the search area? - Anastacio15 commented on September 19th 19 at 13:31
Here seriously, Sergei, you can at least formulate a Monte Carlo. It's not just a coin toss. But it is much on witty. - Vida_Feen commented on September 19th 19 at 13:34
If you specifically need to include the tag I sympathize with you. - jaren.Rutherford commented on September 19th 19 at 13:37
September 19th 19 at 13:24
If the target function is smooth - gradient descent. If not - genetic algorithms.
Feature-it is smooth, but too many local minima. Gradient descent in combination with Montecarlo to select starting points? Or to smooth out outrageously to local minima become less? - earnestine.Smitham72 commented on September 19th 19 at 13:27
: Depends what optimised. The experience of applying the method of inverse iterations in image processing is the choice of the initial approximation does not affect significantly the quality of the result affects only the convergence rate. As far as I know local extrema helps fight annealing method, well, or very bosyatsky - found optimum - "stepped aside" and ran the optimization again. If the result is not better left to the previous option. If better - continue to look. - Araceli.Gorczany commented on September 19th 19 at 13:30
: If you take the image, let it be the combination of images for stitching panoramas. Without searching for feature points only on the global convergence of the fragment :) - Anastacio15 commented on September 19th 19 at 13:33

Find more questions by tags AlgorithmsPHP