Machine learning in trading: theory, models, practice and algo-trading - page 2658

New comment
 
Maxim Dmitrievsky #:
Relates them differently to targets, in the case of transformations. There may be different results. Clustering, in fact, should not give anything :) of course I can show off with clustering of classes, but so far I don't get anything sensible. That is, I take and correct the original zataset at the expense of clusters, supposedly the examples will be better distributed by classes. But it's in the current moment, and it's still hard to predict the appearance of a cluster

I think it is not difficult, but impossible to predict. Identify the current states, categorise them into classes and track their appearance in real life, that's the maximum you can logically do.

 
Valeriy Yastremskiy #:

I think it's not difficult, but impossible to predict. Identify the current states, categorise them into classes and track their occurrence in real life, that's the maximum you can logically do.

Well, it's like taking increments and dividing them by >,<0, and then puzzling over how to write a TS on it :)
 
Valeriy Yastremskiy #:

there's nothing on the graph but increments and time. And derivatives don't give anything new

No way)))))

What you described is the first building blocks of complex structures.

It is the same as saying that there is nothing on the screen except pixels, and derivatives give nothing, but it is not true....

By the way, you forgot to add the price itself
 
mytarmailS #:
No way)))))

What you described is the first building blocks of complex structures.

It is the same as to say that there is nothing on the screen but pixels, and derivatives give nothing, but it is not true....

By the way, you forgot to add the price itself.
I misspoke, they do not give predictive data, but an estimation of the state of course, and a probabilistic estimation of stationarity apparently can be obtained. That's something.
 
Valeriy Yastremskiy #:
I misspoke, they don't give predictive data, but an estimate of the state of course, and a probabilistic estimate of stationarity apparently can be obtained. That's something.
Why not, imagine how much information you get from ordinary pixels every day.
 
mytarmailS #:
Yeah why not, imagine how much information you get out of regular pixels every day
Good comparison )
 
Maxim Dmitrievsky #:
Good comparison )
I agree, good comparison, but with a small clarification, the previous set of pixels makes it almost impossible to predict what the next set of pixels will be like. Or if we go back to our rams, on a piece of the previous history we can perfectly do a lot of different analyses-justifications, but it practically does not give us a chance to predict what the next piece of history will be like.
 
If someone could learn to predict at the opening of each next minute, whether the price will be higher or lower in 240 minutes (i.e. the colour of a 4-hour candle), with a probability of at least 55/45 - that would be alpha.
 
sibirqk #:
I agree, good comparison, but with a small clarification, the previous set of pixels makes it almost impossible to predict what the next set of pixels will be .

You don't know what you're talking about.

When you play a racing game, don't you predict what will happen next? Don't you predict that the turn in the next frame will be closer than it is now, but you have to turn or brake now or else boom.... (prediction)

And all this by pixels, from which our brain builds models, and from the models other models, and those other models ... and so many times, and then we realise that there is a road, turns, speed, steering wheel, etc....


So the problem is not in pixels, or returns, but in our primitive models that we build for the market, the maximum of these models is to multiply pixel by pixel and look at it in a sliding window of 5-10 pixels ) that's the whole tale. That is, above the first level of abstraction model does not rise, and such levels may need 1000 ...


So don't scold returns or pixels, you need to think more with your head, and of course you need knowledge....

 
When you turn off the gas on sizzling cutlets, why do they keep sizzling. Because of inertia. )))) and that's the correct answer.
1...265126522653265426552656265726582659266026612662266326642665...3674
New comment