Uniform concentration inequality for ergodic diffusion processes observed at discrete times (Q1761486): Difference between revisions

The authors consider the process \(\;(y_t)_{t \geq 0}\;\) governed by the stochastic differential equation \(\;dy_t\) \(= S(y_t)dt\) \(+\sigma (y_t)dW_t,\;\) where \(\;(W_t, \mathcal{F}_t)_{t \geq 0}\;\) is a standard Wiener process and \(\;y_0\;\) is an initial condition. Concentration inequalities for the deviation in the ergodic theorem for this process given discrete time observations are proved for a smooth function and an indicator function. A concentration inequality is also proved for continuous observations and the simultaneous geometric ergodicity property for the \(\;(y_t)_{t \geq 0}\;\) process is also proved. The nonparametric point-wise estimation of the drift coefficient given discrete time observations is given as an application.