Machine learning in trading: theory, models, practice and algo-trading - page 2717

New comment
 
Valeriy Yastremskiy properties of an object are often consequences of some causes - actions - events.

But, the name of an object does not usually carry its properties. Well, unless the names are not ranked by properties. After all, the object is described by its properties. A price series is the only object that can be primary described and fixed in the form of ticks. Predictors, clusters, quanta, etc. are terms inside some algorithms of processing this data.

In general, the data is only tick prices, time or tick sequence number, and further filters or averaging are data. The result of processing this data are the same prices, time or serial number of a tick or bar, or their sequence.

Well, if it is a significant sequence, we can call it an event)))

Although it reminds me of the story of the machete. When I saw it, I didn't know it was a machete, and I called it a long axe and for several years I was sure that it was a long axe)))) until I learnt that it is a machete in other countries))))

A meaningful sequence is not an event, they are different concepts. The former is not even a concept, most likely, since there is no definition.

Accordingly, it is impossible to think a meaningful sequence.

There is an elder tree in the garden and an uncle in Kiev

In everyday life, it's okay, but not when you're going to research something.

You should have written "I suppose that I collect events and wrap them in leaves, dear experts, what do you think about it?", you would have been asked to give definitions, found logical errors in them or confirmed the presence of a logical connection.
 
Maxim Dmitrievsky #:
A meaningful sequence is not an event, they are different concepts. The first one is not even a concept, most likely, since there is no definition.

Accordingly, it is impossible to think of a meaningful sequence.

There is an elder tree in the garden, and an uncle in Kiev

You should have written "I assume that I collect events and wrap them in leaves, dear connoisseurs, what do you think about this?" would have asked for definitions, would have found logical errors in them or confirmed the presence of a logical connection.

Undoubtedly, the strictness of definitions reduces the number of errors, logical and other in research)))))) But I don't think it is necessary to be categorical to those who suffer from other definitions)))))) In general, Alexei has the same data and the same tools, and the output is the same type of data)))))

Well inside he is different maybe religion, but it does not change anything)))

 
Valeriy Yastremskiy #:

Undoubtedly, the strictness of definitions reduces the number of errors, logical and otherwise, in research))))) But I don't think it's necessary to be categorical to those who suffer from other definitions))))) In general, Alexey has the same data and the same tools, and the output is the same type of data)))))

Well inside he is different maybe religion, but it does not change anything)))

Purely logically, I did not even see the connection of real events with his indicator rules, so this part of the theory can be thrown out. What is further there is not clear, what then the research is based on.

Okay, this is IMHO, based on logic. I'm not interested in digging any further.

My pressure comes from the fact that the two of them are pushing this event-based model and denigrating the logical approach of time series classification. I tried to find out and realised that the problem is in the head.
 
Maxim Dmitrievsky #:
Purely logically, I did not even see any connection between real events and his indicator rules, so this part of the theory can be thrown out. It is not clear what is further there, what the research is based on.

Okay, this is IMHO, based on logic. It is no longer interesting to dig further.

My pressure comes from the fact that the two of them are pushing this event-based model and denigrating the logical approach of time series classification. I tried to find out and realised that the problem is in the head.

Well, it can't be without an event formalised series in time)))) of course the pre-trend price behaviour is not an event, but a person wanted to designate it so much, it is not just an ordinary series, but an event in the series))))) Of course this is not the fastest way of research, to make your own world, before exploring the boundaries of the studied or at least familiarise yourself with the terminology).

 
Valeriy Yastremskiy #:

Well, it can't be without an event formalised series in time)))) of course, the pre-trend price behaviour is not an event, but a person wanted to designate it so much, it is not just an ordinary series, but an event in the series))))) Of course this is not the fastest way of research, to make your own world, before exploring the boundaries of the studied or at least familiarise yourself with the terminology).

So if you just remove these meaningless bloopers, you're left with the usual MO on time series

Okay, I get it, I get it with this creation.
 

Continued personalities will result in one week bans.

 
Smoothly moved from machine learning to market theory (and on to personalities :)).
 
Vladimir Perervenko market theory (and to personalities :)).
And is it possible to create a successful algorithm without building a more or less working market theory or a market model or a model of the local market situation?
 
mytarmailS #:
Is it possible to create a successful algorithm without building a more or less working market theory or a market model or a model of the local market situation?
One of the most sensible thoughts in the last hundred pages. Imho, of course.
 
mytarmailS #:
Is it possible to create a successful algorithm without building a more or less working market theory or a market model or a model of the local market situation?
Can you handle macroeconomics?) pull out from under the wheels of Economics?
1...271027112712271327142715271627172718271927202721272227232724...3512
New comment
Reason: