Machine learning in trading: theory, models, practice and algo-trading - page 2139
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 don't set a goal in the form of creating a market model, I only set a goal of profiting from price movement...
The truest Goal! And for me, the income is the most correct criterion for evaluating models/approaches, anything, a stable income.
it's a simple logical comparison of everything to everyone else, it's primitive and stupid
Primitive because you can describe it more accurately.
it's dumb because you can't do the inverse.
It's just maraic. So far, I have described pure bars trends according to Williams, from max mines of the first order, from max mines of the second order and from max mines of the third order. Trend of the second order is when the trend does not change its direction after extrema of the second order appear.
Those are also separate states - narrowing and widening of a corridor defined by speeds between extrema of the same order. There is also a JumpUp and a bar fence, when lows of the first order maximums alternate after one bar in the next one. The problem with it is that it seems to be easy to detect, but the data is too different. Generally it is when several bars are like a fence, i.e. it is thicker than a normal corridor.
A trend is also defined by the ratio of the body spread (corridor) to the minimum maximum, though it is not very correct. It may be a thin flat corridor and its length may be determined only by the time of existence, probably, it will be necessary to count the number of bars. It is really complicated here, points and time... I do not know yet. Apparently, if the width of the corridor is normal for a range, it is a flat, and if it is wider, it is a fence.
So, such an interesting world appears on 132 bars))). I plan the logic of channel penetration depending on the state of series and previous states.
In your direction, personal messages are closed, write yourself.
They are open to friends.
I am planning the logic of the channel breakthrough depending on the state of the series and the previous states.) I plan the logic of the channel penetration, depending on the state of the series and the previous states.
When the channel is broken through, do you trade inside or outside the channel?
When you break the channel, is the trade to the inside of the channel or to the outside?
Outside. Inside is more difficult. Although it all depends on the condition. It's just that the algorithm to the outside is obvious. On the inside, we should think)).
Just maraing. So far, I have described the Williams track.........
i don't get it)
I don't get it.)
Well, about the coding of the state of BP was a question. I wanted to use numbers at first, but I realized I would not be able to memorize everything, so I decided to represent the stationary segments in abbreviated text. Short term intermediate extrema? I have the first and the second and the third orders. I will use the code in the comments. It is much more informative with Williams than with WP.
Well, there was a question about the coding of the BP state. At first I wanted to use numbers, but I realized I would not be able to remember everything, so I decided to mark the stationary sections in abbreviated text. Short term intermediate extrema? I have the first and the second and the third orders. I will use the code in the comments. And Williams is more informative than ZZ.
I'm not a friend, I'm talking about something else, Williams is unlikely to work here.
No, I'm not. I was talking about something else.
I'm not talking about Williams, he's an example, I got carried away. I'm talking about coding. I guess I didn't get it all the way through. I mean how to make the zigzag data of different scale, in relative units of data, maybe it will help.
I'm not talking about Williams, he's an example, he got carried away. I'm talking about coding. Lan, I guess I did not understand it to the end. About how to make the zigzag data of different scales, in relative units to prepare the data, maybe and will help.
If I wanted to check my strategy with trend lines, then I'd better convert everything to symbols, so what's the point? )))
In addition, I know how to do it, I was just curious to hear more options ...
Well, I do not know... I hastily finished the first draft so to speak algorithm, the generalization is not very good but sometimes something does guess.
If someone's going to look into it, we can discuss how to represent data so that AMO understands something, although I haven't even tried it directly without normalization.