Machine learning in trading: theory, models, practice and algo-trading - page 1399
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 looked at the map and immediately saw everything. Like a genius Zhukov. (Exactly megalomania.))
Is it normal for you to work with a model with an 80% error?
this is more like mania).Is it normal for you to work with a model that has an 80% error rate?
This is more like mania.)Again he hangs labels. What to do with our Guru? There is no way out of the role).
Do you see any difference between the distribution and the error?
Again he hangs labels. What to do with our Guru? There's no way out of the role).
Do you even see the difference between a distribution and an error?
Between the distribution of the error and the error that is distributed?
what labels, I'm just wonderingbetween the error distribution and the error that is distributed?
what labels, I'm just wonderingWhy am I convincing you? Why do I need this, about the simplest things for an hour. It's useless anyway. You don't understand the results and you don't have to. Stay with your opinion.
One request, do not talk about what you do not understand. Talk only about what your concepts allow you to do. So: not knowing the laws of the Iroquois language, can you make a judgment on this subject that would not be groundless and stupid? (с)
1) Taking into account the real spread can make ellipses less beautiful. It is especially important for small time.
2) It is not only the ratio of profits and losses that matters, but also how they are mixed in time. Unsteadiness always leads to accumulation of losses in a series and large drawdowns. You can't see this on ellipses.
1) Taking into account the real spread can make ellipses less beautiful. This is especially important for small time.
2) It is not only the ratio of profits and losses that matters, but also how they are interspersed over time. Unsteadiness always leads to accumulation of losses in a series and large drawdowns. You can't see this on ellipses.
If you read my post, my trading robot has futures everywhere. The spread is 1 point. The commission is 2-3 pips, and 1-1.5 pips for intraday. I should just spit and delete it.)).
2. In the posts there is a graph of the prototype of the TS for the test of 3.5 months. (or read my blog, I copied some of the posts there). Everything is confirmed. Well, and such a real system, of course, nobody is going to do, the TS is only for confirmation of what we have already seen from the previous one.
3. And the price and even the output of the LPF I have not predicted. Only the unknown center of the unknown distribution of the LPF output in 5 min. I do not know where they, the price and the LPF (like MA(10-12), will take away in 5 m.)
Well, read the posts themselves, all this is already written there. I can not repeat 100 times the same thing.
ZS. I thought, the formulation of the problem is not new, and solved by mathematicians everywhere in the early 40's. (And now it is not outdated either).
1.If you read the posts themselves, I have futures everywhere. Spread - 1 p. Commission 2-3 p, for intrade 1-1.5 p. Nothing at all. I should just spit and delete it.)).
2. In the post there is a graph of a prototype of the TS on the test for 3.5 months. (or read my blog, I copied some of the posts there). Everything is confirmed. Well, and such a real system, of course, nobody is going to do, the TS is only for confirmation of what we have already seen from the previous one.
3. And the price and even the output of the LPF I have not predicted. Only the unknown center of the unknown distribution of the LPF output in 5 min. I do not know where they, the price and the LPF (like MA(10-12), will take away in 5 m.)
Well, read the posts themselves, all this is already written there. I can not repeat 100 times the same thing.
If you take a relatively flat chunk of the chart (trades from about 75 to 175), the gain per trade will be about 7, so the spread and commission will eat up about half. It's really not worth the trade.
The drawdown at the end reminds of one fellow who demands the key from everyone.)
And the sag at the end reminded me of a fellow who demands a key from everyone)
That man is me! I am that man, Oleksiy! Did you find the key? Show it to me! God will thank you.
If you take a relatively flat chunk of the chart (trades from about 75 to 175), the gain per trade will be about 7, so the spread and commission will eat up about half. It's really not worth the trade.
And the drawdown at the end reminded me of a fellow who demands a key from everyone)
Well, this is not a real system). And it is quite possible to trade, with stops, trailing, and other attributes of transaction support. And who says that you have to close in 5 minutes? )) And you can open not only by this prediction, but by additional factors.
You all confuse the working TS with the results of the experiment, which solves only the problem and nothing more.
It is interesting that no one asked - what problem and what kind of mathematicians solved it? Actually I wrote about it several times and gave references, and not a single soul asked.) I won't repeat myself.) Not by force.
No kidding - what kind of discontinuity indicators do you have experience with - Hearst, H-volatility, entropy, autocorrelation, something else? Can you describe this experience in your article?