Algorithm Optimisation Championship. - page 68

New comment
 
Реter Konow:

1. So an evolutionary algorithm is also called a complete enumeration of all values?

2. That's exactly the point I made in this paragraph. You repeated the meaning of what I said.

3. About the painfulness - my subjective impression, but you refused to consider range and pitch changes as a matter of principle.

4. Right, that's how the FF is presented. But the optimum value may not be the maximum. You haven't mentioned this once. But you could have...

5. No expert in his field would give analogies for understanding that could be misleading. This is obvious. Your analogies were from the realm of science fiction.

6. At no point have you clearly articulated what FF is. Your explanations have contributed to such a wide range of perceptions in such a wide field of imaginative play that one could not help but question your own understanding of what you are talking about.

All of the above explains my doubts about your competence on the subject, but it doesn't mean that you are really incompetent. Perhaps you are just that kind of expert.

Regarding the fact that you do not have to explain anything to others and do them a favour, I agree with you. I do.

1. A complete overshoot is not optimization. It is a complete overshoot.

2. The point is different. You are suggesting imperfection of my algorithm. And I'm talking about the impossibility of judging perfection or imperfection without organizing a competition.

3. What the hell is this?! I didn't say no, I said the range can be anything, but not all participants will agree to post their algorithms without a limit, this is a protection of your intellectual rights.

4. How do you say "best" by mathematicians? - only by presenting a function where the best value is the maximum!

5. You asked for an explanation. I explained. You said I didn't explain correctly on the basis that you didn't understand my explanation. Draw your conclusions. Your lack of understanding is not my problem.

6. I will repeat it again. Tomorrow I will present the function as a compiled *.ex5 library. This is the FF max to be found. Search, compete, who will show the result with the value more than others, he will win. Results should be given as FF value and list of parameters obtained. What else needs to be said? How else to make it clearer? Find the maximum! - two simple words which cannot be interpreted otherwise than "Find the maximum!"

 
Реter Konow:

...

Andrey Dik:

.....Ok. I'll ask you a puzzling question, your answer to which will give a complete idea of the level of training: "Which algorithms will get an advantage and why for FF(f1(x1,y1)+...+ f250(x250,y250)?"

Got an answer to that question?

OK, I'll give you a hint in the form of a leading question: "Why does the algorithm on the GCF give such excellent results on 2 parameters and why can it never, for the foreseeable time, with an unlimited number of FF runs, get that close to the maximum if the parameters are, say, 500?"

Or try to use HSS to find such a combination of letters, and there are 33 variants of letters as you know, such a phrase:

Suppose the man has an algorithm for identifying the useful component in the shochastic, non-stationary, noisy, generally close to the markets series. The algorithm requires computational power and time. There is no alternative to it or any other way to solve the problem. The Championship is needed at least to know if there are other, more rational solutions, and as a consequence, whether it is worth digging further. It remains only to properly provide the conditions and data to extract similar information from the algorithm.

There are exactly 500 characters. Now consider which algorithm would do the job best and at what cost.

 
ILNUR777:
Those who have one really don't like tricksters. Especially when they're fiddling around like a tiger in a frying pan.
Like-I have an algorithm, but I'm not going to fall apart, but I want to see the results of similar algorithms, even if without disclosing them. ))))
A number of people have developments close to the Nobel, from which stage they need championships. And the championships between well-known results on applicability, who the fuck need them, to compete in coding if only. But it's not really the coding that rules, it's the approach.
In this regard it's more for you to prove your claims about the quality of your approach, and not those whom you invited to the tournament and said that your algorithm is much better. Order of magnitude is higher not in speed, but in the speed/result combination.
I didn't claim my algorithm was better. I never did.
Let me tell you a story. Some time ago, my Expert Advisor was lying in the Market. It was a tool that took the optimization results of the Expert Advisor in the standard tester and could sort out the optimization results by criteria. This way you can find out which criteria influence the successful forward movement and apply those criteria during optimization. The product had a symbolic price of $10. And for all the time the product was being shown there hadn't been a single sale! More than a year passed and I withdrew the product from sale. People don't want to pay even 10 quid if they have to think, they'd rather pay 3000 for an expert black box.
 
Andrey Dik:
People don't want to pay even 10 quid if they have to think, they'd rather pay 3000 for an expert black box.

Naturally. People would rather risk $3000 for a chance to get a working money-printing machine than invest $10 in a tool that helps them realise that such a machine does not exist.

And it's not easy to sell a product even for $10. You shouldn't have hoped it would sell itself.

PS: the championship leak counts!

 
Post your problem, Andrew, and see how it can be solved.
 
Реter Konow:
Put your problem out there, Andrei. Let's see how you can solve it.
How will you look into it? Are you going to decompile *.ex5?
 
Andrey Khatimlianskii:

Naturally. People would rather risk $3000 for the chance of getting a working money-printing machine than invest $10 in a tool that makes them realise that such a machine does not exist.

And it's not easy to sell a product even for $10. You shouldn't have hoped it would sell itself.

PS: the championship leak counts!

I don't understand why you're gloating, namesake...
 
Andrey Dik:
How do you view it? Are you going to decompile *.ex5?

You started the Championship in order to check whether your algorithm of optimization is the best one among other algorithms.

Do you want to know - have you reached perfection? Aren't you the strongest in this area?

I want to try to help you find out.

You could find this out with a championship, but organizational issues get in the way.

Let's keep it simple: Give me the task that your algorithm does best. Show the result. Prove it's true.

I, in turn, will try to do better. It will take me some time to study the subject additionally, but then I will state whether I did better or not.

In any case, by and large, you will achieve the main goal, and all your work to organize the championship will not be 100% useless ...


P.S. Just think that the challenge that you've given to one person will be taken by another person.

 
Andrey Dik:

Ah in that sense... Well, I said I'd post the compiled FF today. Decide. Compete. Since the algorithms are now on hand and don't need to be posted publicly, we can remove restrictions on parameter boundary conditions (I think that's what you wanted). So the tasks will be very difficult. The first task will be this: who is the first who can compose a meaningful text of 500 characters (degree of compliance with the initial text is calculated within FF and given as a result - a larger number is better), and he wins.

ZS. What about my questions? Will they remain unanswered?

1. I have accepted your challenge.

2. Let's have a competition. You and me. The task of composing a meaningful text has nothing to do with our topic. "Meaningful text" is subjective. To me, what you just wrote is nonsense, and to you, it makes sense.

3. You know what I'm talking about, I'm offering a fair contest.

4. Let the problem be in an open form. So that everyone may try to solve it better than you.

Then you will have a championship...

 
Реter Konow:

1. I accepted your challenge.

2. Let's have a competition. You and me. The task of composing a meaningful text has nothing to do with our topic. "Meaningful text" is subjective. To me, what you just wrote is nonsense, and to you, it makes sense.

3. You know what I'm talking about - I'm offering a fair contest.

4. Let the problem be in an open form. So that everyone may try to solve it better than you.

There will be a championship for you then...

Andrey Dik:

Oh, in that sense... I said I'll post the compiled FF today. Decide. Compete. Since I now have the algorithms in hand and don't need to make them available to the public, I can remove restrictions on the parameter boundary conditions (I think that's what you wanted). So the tasks will be very difficult. The first task will be this: who is the first who can compose a meaningful text of 500 characters (degree of compliance with the initial text is calculated within FF and given as a result - a larger number is better), and he wins.

ZS. What about my questions? Will they remain unanswered?

Do you see bold? A bigger value is better. Whoever finds the biggest number is the winner.

What do you mean, "The problem must be open-ended?"

1...616263646566676869707172737475...133
New comment
Reason: