View Single Post
  #4   Spotlight this post!  
Unread 03-14-2016, 10:38 AM
Ether's Avatar
Ether Ether is offline
systems engineer (retired)
no team
 
Join Date: Nov 2009
Rookie Year: 1969
Location: US
Posts: 7,997
Ether has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond reputeEther has a reputation beyond repute
Re: Shooter speed calculation from vision tracking

Quote:
Originally Posted by Alon View Post
so a piecewise linear function will be more accurate than a polynomial regression?
Not as a general proposition, no. It depends on your empirical data.

For data that is shaped such that it is not well-modeled by a polynomial of reasonable degree, a piecewise-linear function might be more accurate.

Also, depending on the processor being used, a piecewise-linear function may execute faster than a high-order polynomial.

Quote:
and physical calculations will be generally more precise than a table?
I don't know what you mean by "physical calculations" so I'll wait till you have clarified before answering that.


The thing you should be focusing on is consistency. If you shoot several times from the same location using boulders of different weight, age, and state of disrepair, do you always score the goal, regardless of the battery's state of charge? If you don't have a consistent shooter, it's fruitless to worry to much about polynomial vs piecewise-linear vs lookup table.