Testing real-time forecasting systems - page 34

New comment
 
grasn >> :

to gpwr

Just out of curiosity, is your prediction conceptually similar or is something else coming out?

Sorry for the delay. My computer got a virus and I had to treat it for two days and then restore system files and Russian language support.

Here is the last prediction.



For the new prediction for the next week I will try to use smoothed prices (by regular SMA according to PRICE_TYPICAL):


And here is the prediction without smoothed prices (as in previous posts):



So, expected to move slightly up and then sharply down.

 
benik >> :

Yes, curiosity is certainly piqued. :) Because I do practically the same thing myself.

But, ok, I won't ask any more questions. All I will say is that I have come to conclusions that are opposite to those of those above. Namely - the higher the timeframe, the worse the chances of winning the market. And by the way, kamal - a former forum member (you probably remember him) - stated the same thing. It goes something like this: "the higher the timeframe, the closer the price is to the martingale".

Of course I do. But I'm talking about the range, and not the price range at all. Remember, point one, it's important. :о)

"The higher the timeframe, the closer the price is to the martingale."

My long-standing price research shows that the most martingale is the ticks, but "higher" is where the "patterns" start to emerge. But that's where the other problem arises - the increase in dispersion.

 
gpwr >> :
...


So, a small move up and then a sharp move down is expected.

Yes, this is the second alternative that the model shows after processing the statistics and the initial state probability density vector of the system (picture below). This move is also possible and quite "strong" in probability terms. For this reason, it is better to wait 1-2 days (this is if trading is reliable or risky :o).


 

I agree about "up and then down"


 
By the way, gentlemen, for some reason we publish only EUR-USD forecasts, what about publications for other pairs, I just have 10 working charts and sometimes my forecasts are much stronger and clearer than euro-bucks, how do you feel about it?
 
NEKSUS_ >> :
By the way gentlemen, for some reason we publish only EUR-USD forecasts, what about the other pairs, I have 10 working charts and the forecasts are sometimes much stronger and clearer than euro-bucks, how do you feel about it?

i think it would be good but it may be a mix of different images, 10 is too much)

 
kosa >> :

>> it'll be pretty good, but it could be a mash-up of different pictures, 10 is too much.)

hey hey, drop me a line when you get a chance, I haven't seen you in a while.

I think it would be pretty bad, but it might be a mess of different pics, 10 is too much :) And about the screens, I don't mean that I want to do all 10 pairs, good predictions (strong, quality signal from all systems I use) at best 4-5 pairs a month, I thought I would do it.

 
Put it out there, it will be very interesting.
 

Here's what I liked best today:

AUD-USD

EUR-JPY

USD-CAD

 
NEKSUS_ >>:
Товарищь Ko1dun, а вы ещё наблюдаете за нашей веткой, я хотел у вас спросить, вы могли бы поделиться с нами, вашим индикатором, ооочень уж мне вашь прогноз с 25-ой страницы понравился

Thanks for the appraisal.

I cannot share the indicator.

I will only share the predictions.

Here's my last forecast.

And here is the new one.

And GBRJPY as well.


1...272829303132333435363738394041...93
New comment
Reason: