Geek Answers Handbook

Optimization of parameters using the AI ​​method

I know that my question is general, but I am new to AI. I have an experiment with some parameters (almost 6 parameters). Each of them is independent, and I want to find the optimal solution for the maximum or minimum output function. However, if I want to do this in a traditional programming technique, it will take a lot of time, since I will use six nested loops.

I just want to know which AI method to use for this problem? Genetic Algorithm? Neural network? Machine learning?

Update

In fact, a problem can have more than one rating function. It will have one function that we need to minimize (cost) and another function that we want to maximize (power). Perhaps other functions can be added. Example: Building a glass window can be done in millions of ways. However, we need the strongest window at the lowest cost. There are many parameters that affect the pressure capacity in a window, for example, glass strength, height and width, window tilt. Obviously, if we go to extreme cases (the largest strong glass with the smallest width and height and zero tilt), the window will be extremely strong. However, the cost of this will be very high.

I want to study the interaction between parameters in a certain range.

+4
source share
3 answers

Unaware of a specific problem, it sounds as if genetic algorithms would be perfect. They were used to optimize parameters and often yielded good results. Personally, I used them for narrow parameter ranges for border detection methods with about 15 variables, and they did a decent job.

Having multiple evaluation functions does not have to be a problem if you code this into the fitness function of a genetic algorithm. I would look for multipurpose optimization using genetic algorithms.

I would start here: Multipurpose Optimization Using Genetic Algorithms: A Textbook

+1
source

, , .

, ... :

value = strength - k*cost


value = strength / (k1 + k2*cost)

, , , , , . , , , , , ?

... , (.. , , ), ( , ).

, , , - :

  • , n ,
  • 1 <= k <= n k , n, .
  • , .
  • 3 , .

, , , .

+1

Some time ago I wrote C ++ code to solve optimization problems using genetic algorithms. Here it is: http://create-technology.blogspot.ro/2015/03/a-genetic-algorithm-for-solving.html

It should be very easy to follow.

0
source

More articles:


All Articles