greatgonzos
New member
- Joined
- Jan 7, 2011
- Messages
- 1
Why does a Pivot-Point moving average (PPMA) calculate the endpoint of a Linear Regression line?
Hi everyone, I'm a noob here and this is my first post. The reason I am posting is because in one of my recent university classes, we studied "time-series forecasting" techniques. We studied moving averages (MAs), as well as regression. I found the subject quite interesting, and I decided to conduct some of my own reasearch using these techniques.
My question:
Why does a "pivot-point moving average" (PPMA) that uses reverse linear wieghts (that begin with positive values for the most recent data, and end with negative values for the oldest data) calculate the endpoint of a linear regression line?
(By the way, "endpoint" = right-most point if graphed on an xy-coordinate system)
I think the PPMA is also called a "least squares moving average" (LSMA) or "endpoint moving average" (EPMA).
First PLEASE view the two attachments, and then see below.
[attachment=1:2ry8mr16]PPMA.jpg[/attachment:2ry8mr16][attachment=0:2ry8mr16]LR.jpg[/attachment:2ry8mr16]
NOW HERE'S THE INTERESTING PART:
IN the attachments, see how the value corresponding to the most recent complete time period (Fri) is "2.746"? This value, 2.746, is right-most Y-value (the endpoint) of the LR line. It is ALSO the EXACT same number that was calculated using the "PPMA" formula above for the same data!!!
I would really like to know the reason for this, and that is my question. It may involve a form of calculus which may be over my head, but I am more than willing to try to learn and give it a shot.
Math classes I've taken: College algebra, trig, finite math, statistics and probabilities (used the Triola book), and pre-calc.
Thank you,
- greatgonzos
Hi everyone, I'm a noob here and this is my first post. The reason I am posting is because in one of my recent university classes, we studied "time-series forecasting" techniques. We studied moving averages (MAs), as well as regression. I found the subject quite interesting, and I decided to conduct some of my own reasearch using these techniques.
My question:
Why does a "pivot-point moving average" (PPMA) that uses reverse linear wieghts (that begin with positive values for the most recent data, and end with negative values for the oldest data) calculate the endpoint of a linear regression line?
(By the way, "endpoint" = right-most point if graphed on an xy-coordinate system)
I think the PPMA is also called a "least squares moving average" (LSMA) or "endpoint moving average" (EPMA).
First PLEASE view the two attachments, and then see below.
[attachment=1:2ry8mr16]PPMA.jpg[/attachment:2ry8mr16][attachment=0:2ry8mr16]LR.jpg[/attachment:2ry8mr16]
NOW HERE'S THE INTERESTING PART:
IN the attachments, see how the value corresponding to the most recent complete time period (Fri) is "2.746"? This value, 2.746, is right-most Y-value (the endpoint) of the LR line. It is ALSO the EXACT same number that was calculated using the "PPMA" formula above for the same data!!!
I would really like to know the reason for this, and that is my question. It may involve a form of calculus which may be over my head, but I am more than willing to try to learn and give it a shot.
Math classes I've taken: College algebra, trig, finite math, statistics and probabilities (used the Triola book), and pre-calc.
Thank you,
- greatgonzos