Machine learning in trading: theory, models, practice and algo-trading - page 1285
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
Don't be so puffed up). You have 5 products (or didn't you write them?) and a bunch of signals (probably from your own EAs), judging by which you are already living off the income from forex.
And I'm still searching and living from a very different perspective.In fact, one advisor works in signals, just using different channels and their settings for basic signal formation, different combinations of filters, and different MM. Yes it's big, and I wrote the logic, but different functions for working with orders, and because of that I also wrote an auxiliary class for MM, for working with files, some tricky indicators, i.e. I didn't write complex code, but I dragged many of my ideas from there to my EA for MO. It's more than a year I haven't improved the EA, although I have some notes on paper.
I described the logic of that EA many times - it's a grid of orders on a floating channel with increasing lot by Fibo and the rest is a question of filtering (for initial entry and further set up) and basic settings. Trading against the trend, but not in a flat (as I understand it) - the idea is to try to find signs of the end of the trend.
Risks there in any case are big, and not even by the logic of the TS itself, but by the need to withdraw / replenish the account, which would normally withstand the drawdown on short intervals. So on January 03, 2019 I did not have time to transfer money to accounts (there were about 50 cent accounts), which needed to be deposited and as a result I lost all the profits for 2018, but it was certainly my fault - January moves can be very strong, I get convinced every year and plan to turn off trading, and again I have stepped on a rake. Apparently it is necessary to legislate a suspension of trade a week before the New Year and for 3 weeks, or maybe all of January.
So from the market I do not live, but only dream about it. And I think the main reason is slow programming - I have no time to check and code all my ideas.
By the way, cool sigs from Alexei, so I guess I'm lucky\not lucky, but still cool
Yes, there is just fatalism, and in the titles says it. The main thing is to withdraw the initial investment, that it would not be a pity, but as soon as I get ready, I have enough money again - a lot of signals opened...
In fact, one advisor works in signals, it just uses different channels and their settings for basic signal formation, different combinations of filters, and different MM. Yes it's big, and I wrote the logic, but different functions for working with orders, and because of that I've written an auxiliary class for MM, for working with files, some tricky indicators, i.e. I didn't write complex code, but I dragged many of my ideas from there into EA for MO. It's more than a year I haven't improved the EA, although I have some notes on paper.
I described the logic of that EA many times - it's a grid of orders on a floating channel with increasing lot by Fibo and the rest is a question of filtering (for initial entry and further set up) and basic settings. Trading against the trend, but not in a flat (as I understand it) - the idea is to try to find signs of the end of the trend.
Risks there in any case are big, and not even by the logic of the TS itself, but by the need to withdraw / replenish the account, which would normally withstand the drawdown on short intervals. So on January 03, 2019 I did not have time to transfer money to accounts (there were about 50 cent accounts), which needed to be deposited and as a result I lost all the profits for 2018, but it was certainly my fault - January moves can be very strong, I was convinced every year and plan to turn off trading, and again I have fallen on a rake. Apparently it is necessary to legislate a suspension of trade in the week before the New Year and for 3 weeks, or maybe all of January.
So from the market I do not live, but only dream about it. And I think the main reason is slow programming - I have no time to check and code all my ideas.
And those signals that survived - look cool!
Big project implemented. One description takes so much space... )
And those signals that survived - look cool!
Yes, that project is many years old, not everything worked out at once, I accelerated in development when I realized that no one would write an EA according to my TOR and had to go deeper into coding myself. The basic TOR is in 2014 and it is in the public domain now, if you can understand my bone language not the programmer.
Thank you for the signal evaluation. But the increment there is misleading, no 1000 percent there obviously, and I think this misrepresentation is driving people away from the service in question.
If it is good, I should check basic settings for correctness at least once a year and, if necessary, re-optimize them, but I don't do it because of lack of time and resources, everything is taken up by the MO lately.
I limited the tree to 1 line:
samples++; if(samples < 20){ then don't divide the node anymore, but leave the leaf}
i.e. there will be at least 20 examples left in the sheet, for representativeness.
That's the whole release you asked for)))
The degree of undertraining, i.e. the number of samples in the leaf can be any 10, 100, 1000 or optimized. In xgboost it is called min_child_weight
Although it could be simpler, right at the input to DFBuildTreeRec, outside the loop, to calculate
samples=idx2-idx1+1;
In general, as always, one problem can be solved in different ways.
Although it could be simpler, right at the input to DFBuildTreeRec, outside the loop to calculate
samples=idx2-idx1+1;
In general, as always, one problem can be solved in different ways.
Does it make sense to bother, does anything improve?
I assure you, the Zigzags are in no way relevant to BP analysis. You can throw out all the articles on this topic.
You are exaggerating mister, then you will have to throw out or correct (what facepalm as redrawing indicators) all articles on MO, both from this forum and dozens of others, including articles ofVladimir Perervenko who is also targeted - ZZ, andVladimir Perervenko is a professional, no one will doubt, look his articles there whole books and everything as scientists. Most likely you are not able to work with ZZ, remember the proverb about the dancer?
You are exaggerating mister, then we will have to throw out or correct (what facepalm as redrawing indicators) all articles on MO, both from this forum and dozens of others, including articles ofVladimir Perervenko who is also targeted - ZZ, andVladimir Perervenko is a professional, no one will doubt, look at his articles there whole books and all like scientists. Most likely you specifically do not know how to work with ZZ, remember the saying about the dancer?
Does Perervenko have any results?! He does not and cannot have them by definition.
I am not going to work with ZZ, because it absolutely contradicts the very meaning of time series, because the concept of "time" is lost.
Perhaps I am too fixated on the theory of stochastic processes, where time is the fundamental variable. Let someone change my mind - show me the account monitoring on the strategy with ZZ.