Machine learning in trading: theory, models, practice and algo-trading - page 454
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
Okay... I'm not mad at you..... I'm just curious,you know,purely theoretically..... Just for the sake of an experiment. I will send my dataset again, it will involve 3 futures, that is almost 9 months of data, you will build a model and give some verdict. Ideally I would like to run your model on my computer, but I do not insist..... Just curious....
So what's up? Shall I put it out?
Here's where the author of HFT is right, in minutes the NS is very good at picking out entry points. With respect to a random sample the probability of a correct entry increases significantly. To trade directly the NS, of course, will not work, but to use it as part of the usual TS on the logic - the prospects, I suppose, are not bad. Here we can simplify logic itself, and the NS, limiting its problems. Not HFT of course, but intraday may well work out.
Here's where the author of HFT is right, in minutes the NS is very good at picking out entry points. With respect to a random sample the probability of a correct entry increases significantly.
By the way the correlation is very weak on the forex on the minutes, 51-53% accuracy in predicting the next candle, it barely allows to pay the spread, telegraphed about zero, with the right MM, but in the long term it is a loss.
On the minutes on the forex, by the way, the dependence is very weak, 51-53% accuracy in predicting the next candle, it barely allows you to pay the spread, telegraphing around zero, with the right MM, but in the long run is lost.
I checked it on stock instruments. Unlike a few years ago, all the movement is 15-20 minutes... And silence.
In forex, yes, the minutes rather do not rule. Although 51-53% exactness is a very good forecast. But the spread really cannot be recouped.
I checked on stock instruments. Unlike a few years ago, all movement there is 15-20 minutes... and silence.
In Forex, yes, the minutes rather do not rule. Although 51-53% exactness is a very good forecast, but the spread is really not recoverable.( But you can try to use it to improve entry.
What the hell...
a very good forecast ?????
Holy shit...
It's not like we're playing coin roulette.) At 50% "accuracy" on futures, that's ~15% profit per month on the trade volume.
By the way, it's a rare machine that makes more than 50% of correct entries. I don't think it's wrong.)
It's not like we're playing coin roulette.) At 50% "accuracy" on futures, that's ~15% profit per month on the trade volume.
By the way, it's rare for a machine to give more than 50% of correct entries. This is what was said at the conference at the RTS.)
This is what I learned at the RTS conference. This is what they say at conferences...
People write all kinds of crap on the fence, too... What kind of nonsense was not voiced at conferences...
Simulate, at least in Excel. It won't take long, but you'll see for yourself that 50% is not bad at all). Than just throwing words around.
By the way, I have a working model, it's night, but in the evening I can show the results. Although, you still can not convince - if I decide what - I'll drink for sure.
Model, at least in Excel. It won't take much time, but you'll see for yourself that 50% is not bad). Than just throwing words around.
By the way, I have a working model, it's night, but by evening I can show the results. Although, you still can not convince - if I've decided something - I'll drink for sure.
I wrote a bot and it had 25% profitable trades, but in the end it was still in profit by the end of the month. It depends on how you look at the percentage of profitable trades, you can make 95% of profitable ones, but the remaining 5% of losing trades will not only eat all the profit, but also bring losses.
I wrote a bot, so there was 25% of profitable trades, but in the end, all the same in the profit at the end of the month. It depends on how you view the percentage of profitable trades, you can make 95% of profitable, but the remaining 5% of losing trades will not only eat up all the profits, but also bring losses.
How to look at it, imho, is obvious. >0 - profitable, <0 - unprofitable. No matter how you look at it.)
Another matter is that the average profit in the transaction, may substantially exceed the average loss, in the unprofitable. Already wrote - not a coin toss.)
In my model, the limit is somewhere around 40%.
I do not know what people are so obsessed with these 50%.) Also I wrote somewhere - in poker the probability is only 1/6, and good players are always in profit.