Talk:Discretization error
 | This article is rated Stub-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
when we define the derivative of f(x) as f'(x) = lim h->0 [f(x h)-f(x)]/h or f'(x) is approximately equal to [f(x h)-f(x)]/h, where h is a finitely small number. The difference between the first formula and this approximation is known as discretization error.
Difference between Discretization Errors and Quantization Errors
editDiscretization Error
Real number has an important property called density property that says that between any two real number there is a another real number .and so on to infinity
if a and b are two real number then there exist another real number c which is equal to .and also there exist another real number d which equal to and so on to infinity.
this is in mathematics but in computation it is different we cannot say that there is a line we theoretically contains an infinite number of steps instead we have a something called grid or lattice or mesh .
Quantization Errors
It is errors arises due the limitation of floating point representation , it is two types truncation error and round off error. —Preceding unsigned comment added by Ahmedlasheen (talk • contribs) 11:14, 1 October 2008 (UTC)
