NavList:

   

Reply

John, you wrote "... don't insert propositions that are absent from the information at hand."

It seems to me you do not heed this principle. In deriving your 'Estimated 
track made good,' you begin by modifying the information at hand - that from 
log and compass.

As I wrote before, this is exactly the point at which I, a traditional small-boat navigator,
part company with you.

I use the data from compass and log to plot a track and a DR. 
Then I look at the distance between DR and LOP to see if it makes sense in 
light of the other information I have.

I am using my other information at the end. You, on the other hand, look to be 
using your other information (information not from log and compass) at the 
beginning, contrary to the Bayesian principle.

Hewitt


Sent from my iPad

On Mar 26, 2013, at 10:41 AM, "John Karl"  wrote:

> Jeremy,
> 
> P.S.  Not only is the exact shape of the probability distribution not 
critical, but it's unknown.  Such a probability distribution implies previous 
frequency knowledge which is unavailable, and in fact,  unknowable.  This is 
the major point of Bayesian principles: don't insert propositions that are 
absent from the information at hand.
> 
> JK
> ----------------------------------------------------------------
> NavList message boards and member settings: www.fer3.com/NavList
> Members may optionally receive posts by email.
> To cancel email delivery, send a message to NoMail[at]fer3.com
> ----------------------------------------------------------------
> 
> : http://fer3.com/arc/m2.aspx?i=123109
> 
> 

   

Reply