Looking for patterns - page 44

 
And then there's the wavelet transform ... :) Very entertaining mathematics. And it's very good at predicting further movement - exactly as you need to predict the trend, but in physics. True there is a problem of boundary effect there, but at one time digging in this subject in attempts to eliminate this effect I even bought one scientific dissertation on the subject. And I should say that I have overcome the boundary effect. At that time I refused wavelets, because there were no modern neural networks, and they took a very long time to train. Maybe I should go back now ... :):):)
 
Aleksei Stepanenko:

And what use is it? I don't understand.

I would say that this MA often becomes just the extremum boundary. And on all timeframes. At night when volatility is low, and often when there are movements, even on M1 this MA almost always becomes a support/resistance level. Well, of course, it depends on someone. I do not insist on it.

There is a parable on the subject:

Once upon a time, an illiterate and unable to read, but very capable man got into politics. And his position made it necessary to read speeches from a sheet. One day someone noticed that he was holding the text upside down. The man wrinkled his nose for a while, but he parried:

- What difference does it make which side of the paper with the text as long as you can read?

 
Aleksei Stepanenko:

Good point. Apparently I didn't finish ... I only started the topic of the mashka to offer my position and discuss it.

My position is as follows:

1) with the help of this dashka you can determine the moment of turn on the history - this is an important point, because here the dashka will show the true reversal

2) at these reversal points and look for the corresponding regularity with the help of various methods - for me it is easier to do it with neuronets, for someone it is easier to parse this place on OHLCV and to compare similar reversals in the past

3) Also determine how far after the reversal formation the price moved in the past

4) based on steps 2 and 3, draw a conclusion and make a decision.

So it goes.

Regards, RomFil

 

First, think about what you are analysing. Candles? What are they? It's fucking chaos, without any logic.

First, you have to present the stream of quotes in a logical, orderly way. And only then analyze it. You can do it with a wand or a stochastic.

secret has written about 100 times here and not only about rendering, render bars, equinoxes (though, in my opinion, they are no better than regular candlesticks), but if only one posted a picture that doesn't have candlesticks in the background.

 
Макс:

but at least one person would post a picture here that doesn't have candles in the background))

Romphil publishes bars. But that hasn't made it any clearer.

 
Vitaliy Maznev:

Romfil publishes bars. But that hasn't made it any clearer.

Bars are no different from candlesticks in terms of informativeness - i.e. they are useless.

 

Just think - price broken down by the hour (or any other timeframe)...

What is there to analyse? How does the price divided by 60 minutes differ from the price of 58 minutes? It is not... And the indicator readings will be different.

It should not be like that.

 
Макс:
Max:

What's the difference between the 60 minute price and the 58 minute price? Nothing... and the indicator readings will be different.

I completely agree. It's bullshit. It's systematic and orderly. And that's what we have to work with. I don't know how anyone got here, but I'm here because I have no other choice. :)

 
Макс:

Just think - price broken down by the hour (or any other timeframe)...

What is there to analyse? Well, this is nonsense... How does the price divided by 60 minutes differ from the price of 58 minutes? It is not... And the indicator readings will be different.

That's not supposed to happen.

I'll answer in your style...

And prove that the Renko is better. Where are the stats? Show me! :):):)

 
Макс:

Just think - price broken down by the hour (or any other timeframe)...

What is there to analyse? Well, this is nonsense... How does the price divided by 60 minutes differ from the price of 58 minutes? It is not... And the indicator readings will be different.

It shouldn't be like that.

The price listening to no one, but strictly following its corridors.

EURUSD_1

You can apply non-standard candlesticks. This is a matter of taste and perversion.

EURUSD_iH1_6

Reason: