![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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
I'm doing some calculations and I'm deducting the probability of the price direction, let's say, within the current wave, as a percentage.
If there are such gentlemen who are interested in this topic, please join us.
I am interested in your opinions concerning the selection of objects and parameters for analysis.
Let's determine the "direction of price movement" and only after that we may talk about probabilities.
We had a rather interesting thread"What is a trend" https://www.mql5.com/ru/forum/214374 , we still haven't come to a common opinion :-)
Define "price direction" and only then can we talk about probabilities.
There was a rather interesting thread"what is a trend" https://www.mql5.com/ru/forum/214374 but we still have not come to a common opinion :-)
Read the thread))). The thread is fun).
They still haven't figured out the difference between a wave and a trend. Come on.
Let's think of "price direction" as the direction of a wave. Rising and falling. A flat is an undefined state.
Do we agree? Let's not speculate on what a wave is. Everyone should know that. We will not return to this issue.
Here is the same graph, but the ratios have already changed over time.
How does this indicator determine "settlement points"? and it would be great if it could show the ratio with higher timeframes - H1 and M5
How does this indicator determine "settlement points"? and it would be great if it showed a correlation with the higher timeframe - H1 and M5
Here's the H1.
"Settlement points" - do you mean in perspective or from history?
I analyse price behaviour on history and give a future forecast on the output.
Analogue to AI without error checking.
From the previous picture we can draw the following conclusion:
At the moment there is a corrective wave upwards.
There is only a 28% probability of further upside. At the moment (not unimportant).
This might make more sense.
From the previous picture we can draw the following conclusion:
At the moment there is a corrective wave upwards.
There is only a 28% probability of further upside.
That might make more sense.
So what are the probabilities based on? Machine learning uses logit functions(logistic regression and softmax layer for NS), which supposedly give event probabilities (e.g. 0.65 to buy, 0.35 to sell)
Read the thread))). The thread is fun).
Still haven't figured out the difference between a wave and a trend. Come on.
Let's think of "price direction" as the direction of a wave. Rising and falling. A flat is an undefined state.
Do we agree? Let's not speculate on what a wave is. Everyone should know that. We will not return to this question.
simply by the definition of probability, "if the price moved more than P points during dT, it means that something is happening, i.e. the probability of reversal or at least a pullback is high", P is determined by the history of price movements in similar conditions (approximately the same time, on the same days, with similar tick volumes).
So based on what do you take probabilities? Machine learning uses logit functions (logistic regression and softmax layer for NS), which supposedly give event probabilities (e.g. to buy 0.65 to sell 0.35).
I am weak in neural networks of course. For me it is too complicated and tense to make complex calculations. I took a simple way, analyzing price step by step from the history. I do not analyze the entire history, but only the area from 0 bar to the selected by the program itself. I have two screens at most. I do not want to explain the entire process. In general terms. Analysis of the trend, waves, averages, etc.
P.S.
Doesn't require training. Noticed that there are far more errors on learning than simple logic.
If we don't define terms, then what is the probability?
Simply, the definition of probability "if the price moved more than P points during dT, it means that something happens, i.e. the probability of reversal or at least a pullback is high", P is determined by the history of price movements in similar conditions (approximately the same time, the same days, with similar tick volumes).
Already explained in post#75
We're not touching the U-turns yet. We will get to reversals if we are patient.
At the moment we are talking about the probability of further moves. That's it gentlemen.
Already explained in post#75
We're not touching the U-turns yet. We will get to reversals if we are patient.
At the moment we are talking about the probability of further moves. That's how it is, gentlemen.
We take the maximum and minimum of the price from the moment when the movement starts, between them we construct a standard deviation channel. The farther from the middle of the channel, the higher the probability of the reversal. :-)