eldorado.tu-dortmund.de/server/api/core/bitstreams/ec420861-78a4-4ffb-aaec-e1f7a6b59674/content
on
ρ1
ρ 2
−0.5
0.0
0.5
−1 0 1
Ljung−Box
ρ1
ρ 2
−0.5
0.0
0.5
−1 0 1
Von−Neumann
ρ1
ρ 2
−0.5
0.0
0.5
−1 0 1
BDS
ρ1
ρ 2
−0.5
0.0
0.5
−1 0 1
Turning−Point
ρ1
ρ 2
−0.5
0.0
0.5
−1 0 1
3−Depth
ρ1
ρ 2
−0.5
0.0 [...] and second half of the observations by −h/2 and h/2, respectively. Thus, the new process is
14
formally dened as
Ỹt = Yt − 1{t ≤ T/2} · h 2 + 1{t > T/2} · h
2 , t = 1, . . . , T,
in which the autoregression [...] time series (Yt)t=−p+1,−p+2,...,0,1,2,...,T given by
Yt = g(θ∗, Yt−p, . . . , Yt−2, Yt−1, Xt−q+1, . . . , Xt−1, Xt) +Wt (1)
1
for t = 1, . . . , T where (Xt)t=−q+2,−q+3,...,0,1,2,...,T is a d-dimensional …
