Genetic-Algorithms | Genetic Algorithms with example | Machine Learning library

I'm writing an genetic algorithm for finding coefficients having given X, Y points. The principle of operation is described on this page - https://towardsdatascience.com/introduction-to-genetic-algorithms-including-example-code-e396e98d8bf3

I have problem, because sometimes after mutation or crossover my double value is NaN.

I've tried do this using byte[] and BitArray, but in both approaches I have the same result.

Converting double <-> BitArray:

I am trying to make work a piece of code in Python, composed of four modules.

When running I get the following error in the module optimizer_Neuroevolution: "line 149, in evolve retain_length = int(len(graded)*self.retain) TypeError: unsupported operand type(s) for *: 'int' and 'type'

The modules are as follow:

First module main_Neuroevolution:

I couldn't figure out the following problem: if I may, I'd like to give you right away an example
Imagine, you work with marketing data and you came up with a good regression model, predicting the "reach" of a certain campaign. All fine and dandy. Data Scientist Job done.
But wait. We can do more.
My question to you is:
Assuming that we have a good model, how can we optimize the input vector ( = marketing campaign) to get the best possible "reach" ( = predictor / goal to optimize)?
I was googling like crazy, but couldn't find any good approach (I am not talking about any hypterparameter optimization). The best approach I found so far is a genetic algorithm... example here and here
Or - a brute force approach - calculate an enormous grid with tons of possible input vectors and then check, which one is the best (straight forward) - but that would be computational expensive.

I would love to hear your opinion on this. Any advice on which topics I should check out?

I am using python-3.x, and I am trying to generate an initial population that contains random real numbers between 0 and 1 where these numbers should be one of the following: 0, 0.33333, 0.666667 or 1

That means the difference between these numbers is 0.33333 (1/3). I tried to modify this code in many ways but their no luck

I'm writing a genetic algorithm to minimize a function. I have two questions, one in regards to selection and the other with regards to crossover and what to do when it doesn't happen.

Here's an outline of what I'm doing:

I am attempting to run the code from this excellent article: Finding solutions using genetic algorithms in F#

Frustratingly, it reports an FS0001 error, as follows:

This expression was expected to have type 'int' but here has type 'unit'

It omplains that the very last line of the code is incorrect:

I'm trying to port this Genetic Algorithm, and I made a recursive function for advancing from one generation to another.

However, as I'm new to recursion in C# (and in general), I evidently bumped into StackOverflowException when there were too many generations (more than around 4500).

In order to solve the problem, I made Generation() return a bool, so when the genetic algorithm reachs max fitness (goal), it returns true. Else it returns Generation().

If it's about to overflow (Generation > 4500), it returns false.

Now In Main(), to keep Generation() running until it returns true, I use a while loop, so it will start over the recursion until it completes.

This is way more efficient than doing Task.Run, so I went for this approach.

Is this good practice? Are there any more elegant ways of preventing StackOverflows without sacrificing performance?

Population.cs:

I have the following dataset (obtained here):