Statistics as a way of looking into the future! - page 7
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
Well, here's what I've got.
there's no slope at all (. Maybe I did something wrong.
This is the result I believe! The predictability of time series like price series is very low. And if we must exclude the probability of algorithm's error in the results obtained by m_a_sim(like peeking into the future), then everything is true here. Prival, I still see the slope of the cloud, calculate its value. Do not be lazy.
m_a_sim, and for what tool is your algorithm sharpened?
bstone wrote(a)>>
My opinion is that it is practically impossible to look into the future with statistics. And using regressions for prediction is like "the train has already left". As the degree of the regression polynomial increases, the end begins to toss from side to side and you get completely inadequate results on which you'll get tired of making trading logic.
Prival, I can still see the slope of the cloud, calculate its magnitude. Don't be lazy.
Thanks for the idea, I only realised this morning why it's 45 degrees. It's always better in the morning. My indicator has two sigmas (models and measurements) which I was not able to calculate, now I understand what to do with them. I'll try to make a "straight line"
*beer*
My opinion is that it is practically impossible to look into the future with statistics. And using regressions for prediction is like "the train has already left". As the degree of regression polynomial increases, the end begins to toss from side to side and you get completely inadequate results, on which you'll get tired of making trading logic.
It all depends. Even though predicting with statistical methods is driving a car with only a rear-view mirror, using statistics for predictions is still useful in cases where you are modelling the law with statistical methods. For example, if the road is straight, and the statmetods are modelling a straight line, then you only need to keep the road in the rear-view mirror exactly behind you, "in a straight line" so to speak. In this particular case you're absolutely right that you can't get acceptable results, but not because the statistics don't allow it, but because the chosen method isn't adequate to the goal at hand.
That's the way it should be.
Thanks again, something to think about
ЛИНЕЙН(E2:E1440;F2:F1440)=0.934964
That's the way it should be.
Thanks again, something to think about
Wow!!!
Sergey, now please don't go "underground" and comment on your result. After all, according to your data, tan=0.9 and it means almost 100% correct prediction of the expected movement... On what TF? If the TF is different from M1, you're all set!
Are you sure you didn't mix up the indexes when drawing the time series? For example, such a picture will be obtained if the muving is shifted by one or several steps relatively to the current value.
P.S. How should it be - "It should be like this" or like that? - Feel the difference.
Wow!!!
Sergey, now please don't go "underground" and comment on your result. After all, according to your data, tan=0.9 and it means almost 100% of correct prediction of the expected movement... On what TF? If the TF is different from M1, you're all good!
It looks like the value of "polynomial" indicator correlates nicely with the price value.
It looks like the "polynomial" indicator values correlate nicely with the price value.
If so, I will be the first to throw away all my works on NS and join Prival as an apprentice! Right now (or nearly now) I'll start to reread the thread about streams in Forex and build Kalman's filter.
The only pity is that I probably won't have to do it. And the reasons, I hope, will become clear soon.
it's minutes, the shorter the forecast horizon, the more accurate it is. The graph is not drawn from Close, but from the "true price", its estimate.
Here is the chart, the red line is the forecast and the white line is the "true price" estimate. It (indicator) does not redraw.
It's not that simple, needs multi-currency analysis at the very least. And for that you need matrix operations. I still can't make an analogue of transposition as in matcad.
Z.I. It's too early to go into battle yet.
it's minutes, the shorter the forecast horizon, the more accurate it is. The graph is not drawn from Close, but from the "true price", from its estimate.
Here is the chart, the red line is the forecast and the white line is the "true price" estimate. It is not redrawn.
I don't understand what I'm supposed to see here. It looks like a trivial AR(1), i.e. if yesterday it was down, it will be down today, if yesterday it was slightly down, it will be slightly down today. Accordingly, the forecast is late with the price reversal and late with the acceleration/deceleration of the price. That is, if there is a reversal on the zero bar now, the forecast will show it only on the next bar.