Market etiquette or good manners in a minefield - page 49
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 use ready-made, though my opinion is not popular...
What do you mean ready-made? mate labs? or what?
Come on. I will have some rest and I will draw the code on Monday. That's what surprises me:
For example, I take my single layer with 24 inputs and run it on the GBPUSD clockwork. If you give it 200 epochs, the results are disappointing, but if you give it only 24 epochs at 18-20... then I doubt that any two-layer will be able to compete with it - profit 24!
But the "squeezer" for mine is either harmful or(more often) useless, as it jumps over local lows with all the same heating.Alas, there can't be any profits on a normal watch. It is enough to look at the different terminals visually and see different bar data to understand that the system is unstable on watches. This is for the future, so as not to be deceived by small timeframes. A confident forecast can only be made from 4 o'clock timeframes starting, ending, strangely enough, in weeks. The month floats, i.e. it is almost impossible to find the exact entry point on this timeframe without huge drawdowns. One should also take into account crosses. And most importantly, we should use a non-linear predictor after all.
I would like to know more about it. I need a simple example. You can always make it more difficult, but you need to understand how the simplest brick works and then build the house out of it.
>> Sergey, specify please. What kind of example do you need? Perceptron, ORO, or something else?
Alas, there can't be any profits on watchframes.
And most importantly, you have to use a non-linear predictor after all.
I agree with the first one, but I'm not going to trade on timeframes at all
About the second one, categorically not if we're talking about a single layer
I agree with the first one, but I'm not going to trade on timeframes at all
About the second - categorically not, if it is a one-layer
If you don't use multiple timeframes, it's almost impossible to trade. From my experience, because the market is often not all about patterns. Sometimes you have to endure drawdowns, so it is better to look at another timeframe to assess your chances in time, even if you are 99% sure that the currency will come back soon. This "back" can even last for a week, like now with the Eurobucks, which is aiming for 1.3830, for the second week and therefore is going upwards with skewness. As for the second one, the master is the boss.
So I take it that the straight lines are what you get by "regressing" the sets that are marked with crosses?
The red line is how the network was trained, the blue line is how it worked after training. Profit has averaged 20.4 points per candlestick on the hourly USDGBP chart. The girl refuses to work on Eurobucks (it is not visible in the picture)
If you don't use multiple timeframes, it is almost impossible to trade. From my experience, because the market is often not all about patterns.
I'm telling you - you don't need timeframes. Precisely because a timeframe is an inadequate way of representing the market, you need several TFs. But even a dozen TFs will not give you the right view. You should read Einstein at your leisure. But this worthy man said, that the time of any system can be measured only in the events of this system and no other way. What does the hand of a chronometer have to do with market processes? It is an obvious fake!
And we cannot rely on the ticks, because every brokerage company calculates ticks the way it wants - luckily in Forex the ticks (volumes) are not taken into account. So draw conclusions.
I'm telling you - you don't need timeframes. Because timeframe is an inadequate way to represent the market, you need several TFs. But even a dozen TFs will not give you the right view. You should read Einstein at your leisure. But this worthy man said, that the time of any system can be measured only in the events of this system and nothing else. What does the hand of a chronometer have to do with market processes? It is an obvious fake!
And we cannot rely on the ticks, because every brokerage company calculates ticks the way it wants - luckily in Forex the ticks (volumes) are not taken into account. So, draw your conclusions.
In short, I use them:) And I have to use crosses as well, as it's more reliable in determining the pivot point. Well, my depot at least feels more alive that way:)
Sergey, please specify. What kind of example do you need? Perceptron, ORO, or something else.
If I knew what kind of example you need, I wouldn't have asked. Something simple on the mathdeck. Preferably with an explanation of what an epoch is, etc. For me, a lot of terms are incomprehensible, so it often slips the meaning of what you do.
I once saw an example in a textbook how a network is trained on a sine wave, something like that. If it's not too much trouble.