JDAS Seminar –
|Wednesday, December 09, 2015 @ 10:00 am EST — Thursday, December 10, 2015 @ 11:00 am EST|
|Title: Indistinguishable states I: perfect model scenario
In this presentation I will discuss the paper by Kevin Judd and Leonard Smith on Indistinguishable states for perfect models (Physica D 151, 125 (2001)). The aim of this paper is to make an accurate forecast of states for nonlinear systems. It has been shown that even in the presence of infinite past observations, uncertainty in the observations make (true) state estimation impossible even for deterministic model. There will be states those are indistinguishable from the true state. In this scenario, (may be) the only way of making any forecast of the (true) state is by measuring the probability density of indistinguishability. I will discuss a formal derivation of the probability density of being indistinguishability. Later in this paper the authors discuss a new method of state estimation by using gradient descent method.
|Export to My Calendar (ics)|