time series predicting the future price
Hello all,
I just finished watching both youtube video of how to forecast prices https://www.youtube.com/watch?v=w0vSSEq2bn0 . However, my question is where can i find the next days forecasted/predicted price in rapidminer. To clarify. lets say my data ends on April 16,2014 I want rapidminer to predict the next days price April 17,2014. How to I achieve this.
thank you
I just finished watching both youtube video of how to forecast prices https://www.youtube.com/watch?v=w0vSSEq2bn0 . However, my question is where can i find the next days forecasted/predicted price in rapidminer. To clarify. lets say my data ends on April 16,2014 I want rapidminer to predict the next days price April 17,2014. How to I achieve this.
thank you
Find more posts tagged with
Sort by:
1 - 4 of
41
I have same question, could somebody tell me please how to forecast a few next days ?
Altair EmployeeI think you need the series extension to do the things tom did there.
~Martin
~Martin
Yes i installed Series extension. Problem is, no matter which operator I use, result table is still in form :
ID; Label; Predicted Label; Variables ........
How to build model which can forecast bigger horizon than 1 ? With result table that looks like:
ID; Predicted Label
n+1 xyz
n+2 xfka
n+3 ....
etc.
With unknow variables for forecasted variables, just with learned model to catch Trend and seasonability during year ?
Dont need to forecast exact values, only direction of further development.
ID; Label; Predicted Label; Variables ........
How to build model which can forecast bigger horizon than 1 ? With result table that looks like:
ID; Predicted Label
n+1 xyz
n+2 xfka
n+3 ....
etc.
With unknow variables for forecasted variables, just with learned model to catch Trend and seasonability during year ?
Dont need to forecast exact values, only direction of further development.
Altair EmployeeHi,
just a short answer because i need to go:
The trick is windowing. You create a table like this
label var1_n-1, var2_n-1, var2_n-1,
then you can do it. Obviously the window can be longer than one and also contain more complex things.
~Martin
just a short answer because i need to go:
The trick is windowing. You create a table like this
label var1_n-1, var2_n-1, var2_n-1,
then you can do it. Obviously the window can be longer than one and also contain more complex things.
~Martin