The market is a controlled dynamic system. - page 410 1...403404405406407408409410411412413414415416417...551 New comment [Deleted] 2019.03.21 18:11 #4091 Renat Akhtyamov: see it's a thank you. ;)it's a concern for your health. Renat Akhtyamov 2019.03.21 18:13 #4092 Олег avtomat: is a concern for your health.ok show me the test stand. what's the noise? [Deleted] 2019.03.21 18:16 #4093 I do it the way I think it should be done. And if someone thinks it should be done differently, then let them doyour experiment. [Deleted] 2019.03.21 18:34 #4094 Renat Akhtyamov: ok Can you show me a test stream. What's the noise?Tester? No, I won't show it, because I don't run it in the tester. And the tester doesn't run multiple TFs at once. aleger 2019.03.21 19:02 #4095 Олег avtomat: About the next trend is too much. You have to control the current state, which of course is a consequence of the previous state.Nothing is too much - the next trend starts directly from within the current trend (from the moment it begins to reverse),for some time is as if in a "latent" state, then it "appears" and is confirmed at the moment of its visualization (or a bit earlier) as the next current and working trend, it is supported by corresponding work orders, and the above sequence of actions repeats.The whole process (in manual mode) from the beginning to the next reversal and the next continuation, interruption or end is clearly worked out by using the slightly modified Zigzag, the F9 key and is accompanied by the standard income report. Why do the Simple From theory to practice Machine learning in trading: [Deleted] 2019.03.21 20:00 #4096 aleger: Nothing too much - the next trend starts directly from within the current trend (from the moment it begins to reverse),for some time seems to be in a "latent" state, then it "appears" and is confirmed at the moment of its visualization (or a bit earlier) as the next current and working trend, it is supported by corresponding work orders, and the above sequence of actions repeats.The whole process (in manual mode) from the beginning to the next reversal and the next continuation, interruption or end is clearly worked out by using the slightly modified Zigzag, the key F9 and is accompanied by a standard report of the income.We have a different understanding of the term 'trend'. Алексей Тарабанов 2019.03.21 20:23 #4097 He didn't give a definition, why don't you? [Deleted] 2019.03.21 20:41 #4098 Алексей Тарабанов: He didn't give a definition, maybe - you will?A common understanding of the trend in experimental techniques: . ss count how many zigzags he can build here, and call them trends. Алексей Тарабанов 2019.03.21 21:01 #4099 Олег avtomat: A common understanding of the trend in experimental techniques: . ss count how many zigzags he can build here, and call them trends.The 3rd degree polynomial was enough here, but what if the 5th degree is needed? [Deleted] 2019.03.21 21:20 #4100 Алексей Тарабанов: The 3rd degree polynomial was enough here, but what if the 5th degree is needed?This is just an illustrative example. Polynomials are not used here at all. There can be a huge number of experimental points. The task is to identify their overall movement. This will be the trend. By defining a trend, we can describe the experimental data as an additive mixture of a trend and a random component: X(t) = trend(t) + noise(t) HOW TO TRADE FOR How much is needed.... Machine learning in trading: 1...403404405406407408409410411412413414415416417...551 New comment
see
it's a thank you.
;)
it's a concern for your health.
is a concern for your health.
ok
show me the test stand. what's the noise?
ok
Can you show me a test stream. What's the noise?
Tester? No, I won't show it, because I don't run it in the tester. And the tester doesn't run multiple TFs at once.
About the next trend is too much.
You have to control the current state, which of course is a consequence of the previous state.
Nothing is too much - the next trend starts directly from within the current trend (from the moment it begins to reverse),
for some time is as if in a "latent" state, then it "appears" and is confirmed at the moment of its visualization
(or a bit earlier) as the next current and working trend, it is supported by corresponding work orders, and the above sequence of actions repeats.
The whole process (in manual mode) from the beginning to the next reversal and
the next continuation, interruption or end is clearly worked out by using the slightly modified
Zigzag, the F9 key and is accompanied by the standard income report.
Nothing too much - the next trend starts directly from within the current trend (from the moment it begins to reverse),
for some time seems to be in a "latent" state, then it "appears" and is confirmed at the moment of its visualization
(or a bit earlier) as the next current and working trend, it is supported by corresponding work orders, and the above sequence of actions repeats.
The whole process (in manual mode) from the beginning to the next reversal and
the next continuation, interruption or end is clearly worked out by using the slightly modified
Zigzag, the key F9 and is accompanied by a standard report of the income.
We have a different understanding of the term 'trend'.
He didn't give a definition, maybe - you will?
A common understanding of the trend in experimental techniques:
.
ss
count how many zigzags he can build here, and call them trends.
A common understanding of the trend in experimental techniques:
.
ss
count how many zigzags he can build here, and call them trends.
The 3rd degree polynomial was enough here, but what if the 5th degree is needed?
The 3rd degree polynomial was enough here, but what if the 5th degree is needed?
This is just an illustrative example. Polynomials are not used here at all.
There can be a huge number of experimental points. The task is to identify their overall movement. This will be the trend.
By defining a trend, we can describe the experimental data as an additive mixture of a trend and a random component: X(t) = trend(t) + noise(t)