AgTalk Home
AgTalk Home
Search Forums | Classifieds (118) | Skins | Language
You are logged in as a guest. ( logon | register )

One more AgLeader 2000 question
View previous thread :: View next thread
   Forums List -> Precision TalkMessage format
 
plowboy
Posted 6/19/2006 19:39 (#20633 - in reply to #20449)
Subject: Re: One more AgLeader 2000 question



Brazilton KS

I've apparently not done a very good job of clarifying what I am trying to say. The problem with putting all the loads in and trying to use them for calibration is that you are going to have a very large amount of data at whatever your "normal" flow rate is, and relatively little data at the extremes of low or high flow. The machine is going to assign similar importance to all the data. The result is going to be a calibration which is very poor at anything other then "average" conditions. It may or may not do well at average conditions. A secondary problem is that big loads tend to have a lot more varibility within the data in the load, which makes the curve fitting not work out very well. You need a load which has a 500 bu per hour flow rate throughout the load, with a weight which is representative, in order for the machine to figure out what it is seeing when it sees a 500 bu per hour flow rate. The data does no good if it is buried in a load with tons of grain harvested at 2500 bu per hour, because there is no way to decern how much grain was harvested during the time when flow was low.

 

The irony of this discussion is that I know Diet well enough to know he understands what I am saying regarding the regression problems far better then I do, so it's kinda hard to figure out why I don't just shut up....which I guess I will do now.



Edited by plowboy 6/19/2006 19:43
Top of the page Bottom of the page


Jump to forum :
Search this forum
Printer friendly version
E-mail a link to this thread

(Delete cookies)