Machine learning in trading: theory, models, practice and algo-trading - page 1802
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
if it's a number system, you can do it without a huge cycle
you can easily get the bit combination 01101101 from 109 without going through all variants.
Please elaborate.
if it's a number system, you can do it without a huge cycle
you can easily get the bit combination 01101101 from 109 without going through all the variants.
Interesting, but I don't understand the fantasy logic.
if it's a number system, you can do it without a huge cycle
you can easily get the combination of bits 01101101 from 109 without going through all variants.
This is only good if you want to try all combinations, and here we need only those which have exactly K elements (K ones).
In any case, the set of combinations is enormous, and it's unlikely that you can calculate anything for each of them. Here already a couple of times it was advised to count only randomly chosen combinations - it seems more reasonable to me.
This will only work if you want to go through all the combinations, and here you need only those with exactly K elements (K units).
In any case, the set of combinations is enormous, and it is unlikely to get anything for each of them. There've already been a couple of times here advising to count only randomly chosen combinations - this seems more reasonable to me.
Apparently, Alexei is not afraid of this. He's fighting to make calculations not for 250 days, but at least for a month...
Can you suggest a forum for Python and machine learning where you can ask nerdy questions?
If there is a sequence, there must be a formula or other quick solution than going through all the points. Overkill is also inefficient for the application.
It's essentially a function with known points...
I think that you can define areas and build a table within their boundaries. Suppose there is a pattern for each 10000 elements, then from this point to count. It is strange that such a problem is not solved.Once upon a time, I was doing something similar. I had to choose the most efficient passes for combinations of parameters with given conditions. For example, I had 12 parameters, each of them had value from 1 to 10, and I had to check only those runs where the sum of all parameters was say, 27. From them to choose the best runs, to introduce some filter, and so on. It is clear that this task could not be solved directly. So, the first step was the following: I started 12 nested loops from 0 to 9, inside I checked the conditions for the combination, if the condition was met I wrote the string of loop indexes into an array. As a result I got an array with half a million of combinations I needed out of a trillion ones. Then I ran through the obtained combinations, chose the best 100k combinations, copied the indexes to another array and messed with it further.
Once upon a time, I did something similar. It was necessary to choose the most effective runs for combinations of parameters with specified conditions. For example, there were 12 parameters, each of them had a value from 1 to 10, and it was necessary to check only those runs, for which the sum of all parameters was, for example, 27. From them to choose the best runs, to introduce some filter, and so on. It is clear that this task could not be solved directly. So, the first step I did was to start 12 nested loops from 0 to 9, inside I would check for the combination, if the condition was met I would write the string of loop indexes into an array. As a result I got an array of half a million combinations I needed, out of a trillion ones. Then I ran through the obtained combinations, chose 100k combinations, copied the indexes to another array and messed with it from there.
Thinning is our everything))
With thinning alone you will give us daily bread
The only thing left is to find the meaning - the correlation between the scales))))
In the application in one file and balance and OHLCV - maybe so will be more convenient.
I had, it turned out, an error check in the indicator, that's why it happened that way - I should deal with indicators separately - eh.
I have no results, the results are worse than the price itself, there is nothing to show, I think we should go to the deal filter.
By the way, thanks for the perfectly prepared dataset, it is nice to get it working the first time.
I'd rather have a look atVladimir Perervenko's comment, he won't advise much, and R is not so scary to study, on the contrary, he's very friendly, and we, R-people, will be glad to meet a new comrade.