Algorithm Optimisation Championship. - page 49
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
You're on the list, I presume?
Andrey Dik
Tag Konow
Igor Volodin
Dmitry Fedoseev
Sergey Chalyshev
Ghenadie Tumco
Alexey Burnakov
Yuri Evseenkov
Vasiliy Sokolov
You're on the list, I take it?
Yeah, that's right.
Yes . I'm not embarrassed to be in last place. I haven't done any optimization. But in principle, I could adapt one of my designs. You're just taking part with a random number generator, aren't you?
Once I modified MT4 standard RNG with uniform distribution of output values into RNG with normal distribution by polar coordinates method. Well I will suppose that required numbers (let it be 500) are distributed according to the normal law (probability is distributed) within the range double from -1.7976931348623158e+308 to +1.7976931348623158e+308s. Well generate some arrays with normally distributed values and different mat expectations. I will send them to black box. I will get unknown function values from the black box and then I will shift matrix expectations and so on.
My model is B, but I can try to convert it to A for the Championship. The model consists of a single arithmetic and takes a couple dozen lines.
I haven't coded on five, so some questions both on five and on procedural questions (like how to pass an array and how to get a function value into your code) may be confusing.
Yuri, you know, looking at your avatar, I can't help but get the feeling that there are two of you.)
Yes . I'm not complex about being in last place. I haven't done any optimisation. But in principle, I could adapt one of my designs. You do allow participation just with random number generator.
Once I modified MT4 standard RNG with uniform distribution of output values into RNG with normal distribution by polar coordinates method. Well I will suppose that required numbers (let it be 500) are distributed according to the normal law (probability is distributed) within the range doublefrom -1.7976931348623158e+308 to +1.7976931348623158e+308s. Well generate some arrays with normally distributed values and different mat expectations. I will send them to black box. I will get unknown function values from the black box and then I will shift matrix expectations and so on.
My model is B, but for the championship I may try to convert it to A. The model consists of a single arithmetic and takes a couple dozen lines.
I haven't coded on five, so some questions both on five and on procedural issues (e.g. how to pass an array and how to get a function value into your code) may be confusing.
Very good!
But, although we won't know what's inside the FF, there are still restrictions for parameters (done for reasons of protection of participants' algorithms, i.e. there should be restrictions for practical use). Parameter values can be from -10.0 to 10.0 in increments of 0.1.
You can save these restrictions into the algorithm or not, it is up to your discretion, but FF will cut off all numbers in this range. So the task is easier than you imagined.
Very good!
But, although we won't know what's inside the FF, there are still limits to the parameters (done for reasons of protection of participants' algorithms, i.e. there must be limits to the practical application). Parameter values can be from -10.0 to 10.0 in increments of 0.1.
You can save these restrictions into the algorithm or not, it is up to you, but FF will truncate all numbers in this range. Therefore, the task is simpler than you imagined.
Is there an upper limit to the number of elements in the transferred array for FF?
Excuse me. When the actual work starts, I'll ask the dumb questions. Never hooked anything up to mql5. I hope dumb questions are allowed?
I seem to have time to show examples today.
We all ask dumb questions (each in our own time) from the perspective of those who know more about the issue. No problem).