neural network and inputs - page 16
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
So, let's have a peeve in a nutshell. How to find out?????
There, behind the highlighted one, there is: How to do it in two words can't tell, so I'm giving you links where you can read more about it.
In two ways you can not, and in three, too, a lot of text and half a quote from Spider, it is necessary? It's better to read the original than my free paraphrase.
I use one of the methods I use when searching for patterns that have nothing in common with the patterns Neo wrote about, I trade them too, but it's different and really better to read on the Spider. Very close analogues of these patterns are Adverse Tactics patterns.
If we start searching for them the main criterion is not profitability and other parameters of the TS, we are not even talking about trading and testing at this stage.
The main thing is the repeatability of the patterns. If we found a pattern on the watch - we moved to a lower timeframe, for example ten minutes, then to the timeframe above, the daily timeframe. This pattern should be everywhere - otherwise we have found a fake one, send it to the basket. Then we check on other symbols (all available), if none of them shows up, we put it to the basket and look for the next one.
How to trade them later is another problem, not all patterns may be traded, but we have to find them to get the whole picture. So far we have only found the building blocks of which the charts are composed. These patterns may contain different number of bars and have different volatility. It is better, if possible, to move away from timeframes and assemble your own charts from ticks.
If we mark the chart with patterns, it will be found out that only volatility changes, the patterns structure is very stable and does not depend on the instrument/timeframe. A pattern on a higher timeframe consists of the same patterns on a lower timeframe and itself is a part of the pattern on an even higher timeframe. You wrote that this cannot be detected using a neural network, but there is no such a thing as a grid. When patterns are found visually they are perfectly programmable, but detectability becomes worse comparing to definition "by eye".
Highlight the most important thing and press the magic buttons ctrl C and ctrl V. There's no problem..... Everyone and appreciate this judgement ))))
In general, I will tell you this - I personally (3 times imho) sometimes use neural networks, as I wrote above, but as an intermediate stage, so to speak, but in my memory nets have not helped me anything really (while other methods gave results). I don't know why, maybe I just use them "in the first preparation" so to speak, without going deeper and not really bothering with the architecture. I.e., as I wrote above, if I have, for example, an idea, I go at it both ways, including with nets, just for order, so that in case of something "not to spill the baby". However, as I said above, not once have they given me anything tangible. At least compared to the range of other methods (you know, regressors, solvers, nearest neighbours, etc.).
I think the main problem is overtraining (the antithesis of robustness). Actually we all are after robustness, whether we realize it or not. Things are very bad in grids - worse than in other algorithms. According to my observations, the simpler the solver, the more chances it has to be robust.
The less of a woman we are, the less of a woman we are...
Would you suggest a PRIZE, or a different solver?
imho, the solver is in the solver's head.
The smaller the woman we are, the smaller we are...
Will you offer a PRIZE, or another solver?
I think the solver is in the mind of the solver.
Let's say a woman is small, we are bigger (any word will do) and she seems bigger to us than she really is))))
Makes sense!
But how neural networks would understand and whether the answer would be correct - I don't know.
Suppose a woman is small, we are bigger (any word will do) and we think she is bigger than she really is))))
Makes sense!
I don't know how the neural networks would understand and whether the answer would be correct.
What do you mean by right?
By the way, I'm a little confused, it should be like this:
"The more of a woman we are, the less of a woman we are. "
What do you mean, right?
By the way, I'm a little confused, it should be like this:
"The more of a woman we are, the less of a woman we are. "
It's nice to have such interlocutors!!!
YAY!
" The more of a woman we are, the less of a woman we are. "
I'll add...
BRAVO!
Highlight the most important thing and press the magic buttons ctrl C and ctrl V. There's no problem..... All and appreciate this judgement ))))
Have you been banned on Spider or is it a principled reluctance to read? Too lazy? You should be, it's worth it.
Okay. A brief and very loose retelling of the point. The key question in building the MTS(finding the pattern) is: Where does the money come from?
That is, who loses the money when your TS makes it. Study groups of people trading (or just changing currencies on FOREX), their trading methods, try to identify on the chart, build patterns, etc. Find weaknesses in their trading and make a profit where they make a loss. Trying to stand in front of a specific group at a strictly defined moment, rather than the "in general" crowd. (For example, we can select a group of Elillots )
MTS is obtained strictly for certain markets / timeframes and works exactly as long as there is a group of people in the market, against whom our TS trades, taking their money.
A very important feature of MTS based on such patterns, its parameters (ratio of profit/loss trades, profit factor) does not change abruptly. It stopped working once, from 85/15 to 50/50. It is not like that, the graph is stable +- very small deviations (a couple of percent), then it begins to decrease monotonically, at some point the decision to freeze the system until better times is made. Often the pattern starts working again.
The main thing is that you can stop the trading long before the deposit runs out. The system is working very well, but the ratio of profitable/lossmaking trades starts decreasing gradually. This is a signal that the traded pattern becomes inadequate to the market (market phase).
Have you been banned from Spider or is it a matter of principle not to get acquainted? Too lazy? You should be, it's worth it.
OK. A brief and very loose retelling of the essence. The key question when building an MTS (finding a pattern): Where does the money come from?
That is, who loses the money when your TS makes it. Study groups of people trading (or just changing currencies on FOREX), their trading methods, try to identify on the chart, build patterns, etc. Find weaknesses in their trading and make a profit where they make a loss. Trying to stand in front of a specific group at a strictly defined moment, rather than the "in general" crowd. (For example, you could highlight a group of Elillots )
MTS is obtained strictly for specific markets/timeframes and works exactly as long as there is a group of people in the market, against whom our TS is trading, taking their money.
A very important feature of MTS based on such patterns, its parameters (ratio of profit/loss trades, profit factor) does not change abruptly. It stopped working once, from 85/15 to 50/50. It is not like that, the graph is stable +- very small deviations (a couple of percent), then it begins to decrease monotonically, at some point the decision to park the system until better times is made. Often the pattern starts working again.
The main thing is that you can stop the trading long before the deposit runs out. The system is working very well, but the ratio of profitable/lossmaking trades will gradually decrease. This is a signal that the traded pattern becomes inadequate to the market (market phase).
How can this group be identified?
There is a price chart:
1. how can you tell when a particular group is trading? Is it trading? How can I identify a specific group of traders from the chart?
2. OK, a group is identified by magical methods - how do you determine the trading method?
It's the repeatability of the patterns that counts. If you've found the pattern on the watch - move to a lower timeframe, for example ten minutes, then to a higher timeframe, the daytime. This pattern should be everywhere - otherwise we have found a fake one, send it to the basket. Then we check on other symbols (all available), if none of them shows up, we put it to the basket and look for the next one.
How to trade them later is another problem, not all patterns may be traded, but we have to find them to get the whole picture. So far we have only found the building blocks of which the charts are composed. These patterns may contain different number of bars and have different volatility. It is better, if possible, to move away from timeframes and collect your own charts from ticks.
If we mark the chart with patterns, it will be found out that only volatility changes, the patterns structure is very stable and does not depend on the instrument/timeframe. A pattern on a higher timeframe consists of the same patterns on a lower timeframe and itself is a part of the pattern on an even higher timeframe. You wrote that this cannot be detected using a neural network, but there is no such a thing as a grid. When patterns are found visually, they are perfectly programmable, although detectability becomes worse comparing to definition "by eye".
There are some very peculiar ones.
As one hitman used to say, stroking his puppy:
-What a freak you are, buddy.
And they lived happily ever after...