Machine learning in trading: theory, models, practice and algo-trading - page 2707
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
New terms are introduced just to organise the accumulated information and to be able to generalise it and move on.
The fact that I invent different things is just that, something works, something doesn't. You will not deny that you have repeated in your article one of my ideas stated here, will you?
You shouted that I'm a fool for picking leaves a couple of years ago, and now mytarmailS is doing the same thing.
I may be picking up something from the public peeps of reason in this thread, but I don't speak disparagingly until I understand or check.
No, you're completely confused. A point is not an event that turns into a trading system. These are all different terms, it is very difficult to perceive the mess
Representation of quotes in the form of time series is just a way to put information into theory with an attempt to use its developments on this data. At the same time, I do not deny that quotes are time-dependent, but the time-series representation concentrates on the cyclicality of events. The market can and should be represented in different ways and use these representations in the same system. So discrete events, in the form of a single point, can also be meaningful in my representation of the market.
Let's imagine a popular trading system - on RSI breakout, usually it is 70/30 levels, the price closing behind one of the levels will be the fact of the event. Why this is significant - statistically strong trend movements have followed this event. Why it can work - because people use it in their TS.
At the same time, an event can have predictors that describe it:
- Whether or not the event has occurred/what kind of event;
- How much distance before the event occurred;
- How much time has elapsed since the event occurred;
- When the event occurred;
- What happened after the event;
- The duration of the event in time/bars;
- And other things related to the event.
At the same time, we may not see the cyclicality of the event over time.
The market, in my mind, is made up of such events and the relationship between them is just as important - it is an additional layer/model that summarises them.
Well, I will add to this example that it is the correct quantisation of RSI oscillator indicators (as an example) that can reveal these predictors in aggregate, as we will need essentially two quanta of all that describe the value in levels <30 and >70, and the rest can be noise and this noise can participate in learning, which is not necessary.
Regarding, similar thoughts and use of ideas - you have to search - I couldn't find it straight away - I remember talking about the method of cascade sampling separation in sequential learning, you said you were reading about similar systems at the time and couldn't comment. Maybe you read something in books at that time - I don't know, but I mean that you shouldn't say that my ideas don't work if you don't understand them.
Presenting quotes as a time series is just a way of putting information into a theory and trying to use its findings on this data. At the same time, I do not deny that quotes depend on time, but the representation in the form of time series concentrates on the cyclicality of events. The market can and should be represented in different ways and use these representations in the same system. Therefore, discrete events, in the form of a single point, can also be meaningful in my representation of the market.
Let's imagine a popular trading system - on RSI breakout, usually it is 70/30 levels, the price closing behind one of the levels will be the fact of the event. Why this is significant - statistically strong trend movements have followed this event. Why it can work - because people use it in their TS.
At the same time the event may have predictors describing it:
- Whether or not the event occurred/what event;
- How much distance to the occurrence of the event;
- How much time has elapsed since the event occurred;
- When the event occurred;
- What happened after the event;
- Duration of the event in time/bars;
- And other things related to the event.
We may not see the cyclicality of the event over time.
The market, as I see it, is made up of such events and the connection between them is just as important - it is an additional layer/model that generalises them.
Well, in this example I will add that it is the correct quantisation of RSI oscillator indicators (as an example) that can reveal these predictors in aggregate, as we need essentially two quanta of all that describe the value in levels <30 and >70, and the rest can be noise and this noise can participate in training, which is not necessary.
Regarding, similar thoughts and use of ideas - you have to search - I couldn't find it straight away - I remember talking about the method of cascade sampling separation in sequential learning, you said you were reading about similar systems at the time and couldn't comment. Maybe you read something in books at that time - I don't know, but my point is that you shouldn't say that my ideas don't work if you don't understand them.
Nightmare
Your gaze slides across the horizon - the information coming into your head comes about a smoothly changing picture over time - is it a time series?
I can see that there is no effort to understand, I know, pity.
Not surprised, though.
Your gaze slides across the horizon - the information coming into your head comes about a smoothly changing picture over time - is it a time series?
I see that there is no effort to understand, I know, pity.
Not surprised, though.
You keep rambling on instead of getting your head together.
It's not for you and your handicrafts to diagnose me here...
It's not for you and your crafts to diagnose me here...
You want co-op, don't you? For that you have to go through the process of reanimation after two years of neglect and learn to at least just think for starters, otherwise by yourself, by yourself.... 🧠
So start thinking, then. Where is constructive criticism - only some rambling insults.
And for criticism you need understanding, and for understanding you need effort, and you have no such goal - behaviour in the style of "I haven't read it in books, so it doesn't exist and can't exist".
Maximka, our speciality.)
first-tier pioneer
So start thinking. Where's constructive criticism, only rambling insults.
And for criticism you need understanding, and for understanding you need effort, and you have no such goal - behaviour in the style of "I have not read in books, so there is no such thing and cannot be".