Machine learning in trading: theory, models, practice and algo-trading - page 312

 
forexman77:

That's for sure. I work 12-14 hours a day. Well, you still need a distraction once in a while. Over the years, I've crooked my spine.

That's true. I didn't crooked my spine, but now I don't trade every day. Betting on autosystems does not work too well either - a small glitch and you may find yourself without any profit, even with a good system. The leverage, however.

In general, trading, when you already know how to do everything, is damn boring and monotonous - you just sit there, staring into the monitor, and there are only two thoughts - buy/sell. You go to lunch - the most interesting thing has already happened without you).

 
Yuriy Asaulenko:

That's for sure. I have not curved my spine, but now I do not trade every day. Betting on autosystems does not work very well either - a small glitch and you can find yourself without any profit, even with a good system. The leverage, however.

In general, trading, when you already know how to do everything, is damn boring and monotonous - you just sit there, staring into the monitor, and there are only two thoughts - buy/sell. If you go to lunch, all the most interesting things have already happened without you.)


I've been reading this thread for a long time, it is all based on formulas, which is somewhat complicated for me. Are there people, who know complex formulas, willing to code their developments in MQL?

I can try to make a code and at the same time to understand the essence of the process.

 

I recommend to watch a short video of HFT-quant, which is one of the world's leading sabja


Of course, there will be sound

Forum on trading, automated trading systems and testing trading strategies

Analysis of important STATISTICAL characteristics of a pattern and choice of trading method by it.

fxsaber, 2017.04.06 08:30

The grail didn't work out - it wasn't a competent enough approach!

This statement became so popular for what reason?

The luck factor, if not the main one, is certainly essential in this case.

 
fxsaber:

I recommend watching a short video of HFT-quantum, which is one of the world leaders in subj.


Of course they will.

The luck factor, if not the main one, is certainly essential in this case.


Mmmda...not much(( The Facebook generation is going to have a hard time with this sort of thing. The data is low-frequency, probably diaries, what kind of "expensive" features are we talking about?


" Sabz leader" )))))))))) Why are you so quick to troll the boy, let him learn)))

 
fxsaber:

I recommend watching a short video of HFT-quantum, which is one of the world leaders in subj.


Of course they will be sound

The luck factor, if not the main one, is certainly essential in this case.


Exactly... This is an excellent demonstration of approach of almost all adepts of MO - whether in articles on mq5 or elsewhere... We've filed there... <a lot of unnecessary terms and formulas>, and the output is something else... the result is that markets are weakly predictable and neural networks can't predict markets... Another way - neural networks are not for the feeble-minded and children :) find something good, please :)
 
Maxim Dmitrievsky:

Exactly... The approach of practically all adepts of IR is perfectly demonstrated here - either in articles on mq5 or elsewhere... We've filed there... <a lot of unnecessary terms and formulas>, and the output is something else... the result is that markets are weakly predictable and neural networks can't predict markets... Well, as an alternative - neural networks are not for the feeble-minded and children should not be engaged "seriously" :) Find something worthwhile please :)


Not almost all of them, but some of them.

Do you have an alternative to MO that you can post here and clearly demonstrate this alternative to MO?

 
He understandsfinancial markets very well because he is an HFT-Quant himself:

"Leader" )))))))))) Why do you have to troll the boy so fast, let him learn)))

That's the thing, the leader.

It's very well written here how the people who are lower in the ranking do it.

Maxim Dmitrievsky:

Exactly... The approach of almost all adepts of MO - whether in articles on mq5 or elsewhere - is perfectly shown here. We've submitted something there... <a lot of unnecessary terms and formulas>, and the output is something else... the result is that markets are poorly predictable and neural networks cannot predict markets... find something good please :) i didn't expect such videos from you

Oddly enough, I understood absolutely everything in that video. When I thought about it, I paused. When I was ready to listen further, I rewound a bit. Everything is laid out in simple language.

I understand that the age confuses, well, you should have seen for yourself, who the speaker is, rather than compared with his age. One of the most titled and respected among the Defense Ministry in Russia and in the world. And not for the reports regalia. The man knows perfectly well what he is doing. And he proves this in practice in public.

All the more so for a great understanding of financial markets, since he himself is a HFT-Quant.

Kaggle: Британские спутниковые снимки. Как мы взяли третье место
Kaggle: Британские спутниковые снимки. Как мы взяли третье место
  • habrahabr.ru
Сразу оговорюсь, что данный текст — это не сухая выжимка основных идей с красивыми графиками и обилием технических терминов (такой текст называется научной статьей и я его обязательно напишу, но потом, когда нам заплатят призовые $20000, а то, не дай бог, начнутся разговоры про лицензию, авторские права и прочее). К моему сожалению, пока...
 
Dimitri:


Not almost all of them, but some of them.

Do you have an alternative to MO that you can post here and clearly demonstrate an alternative to MO?


What's the point of asking.

I'd say not some, but all ... all of them put whatever into neural networks and call it "predictors", I haven't seen other information yet

 
Maxim Dmitrievsky:


What's the point?

I would not say some, but all ... all of them put not understand what in neural networks and call it "predictors", I have not seen other information yet


If someone puts something into a neural network and it gives you a good result, you will not see and will not be shown.

You need such a cow for yourself.

Is that logical?

 
Dmitry:


If someone shoves something into a neural network and the network gives a good result - you don't and won't be shown this.

You need such a cow for yourself.

Is it logical?


In general, it is not logical, but they show it and take it to the DU... and those who can not take it elsewhere :)
Reason: