20
Denis Belomestny and Markus Reiβ
6.5. Upper bound for μ
The assertion follows as soon as the following Gs (R, σmax)-uniform risk
bound for general U holds:
E h ∞ ∣μ(x) - μ(x)∣2 dxi . U-2s + E(Tσ2U2)U5ε2 + E(2Tσ2maxU2)U9ε4.
-∞
(44)
The bias in estimating μ due to the cutoff at U can be estimated by
1 F μ ( u )(1
-∞
-1[-U,U])|2 du 6 U-2s
∞ |u|2s∣Fμ(u)∣2 du = U-2skμ(s)k2 2.
-∞
(45)
The variance term can be split up according to the different risk contribu-
tions. For u ∈ [-U, U] we obtain
E[IF(μ - μ)(u)∣2]
6 4 E[∣ψ(u) — ψ(u)) 12] + 4(u2 + 1)2 E[∣σ2 — σ212]
+ 4( u2 + 1) E[ ∣γ — γ∣2 ] + 4 E[ ∣λ — λ∣2 ]
. E[∣L(u) 12] + E[∣R(u) 12] + U4 E[∣σ2 — σ212] + U2 E[∣^ — γ∣2] + E[∣λ — λ∣2]
. E[|L(u)|2] +E[|R(u)|2]+U-2(s+1) + E(Tσ2U2)U3ε2 +E(Tσm2axU2)2U8ε4.
In analogy to the previous estimates for σ2 we find
E[∣L(u)12] 6∖φτ(u
. eTσ2u
2u
i ) ∣-2( u4 + u 2)( kF ( O — Oi ) k ∞ + Var[ FO)( u )])
^( ∆4 + ∆2 kJk22 ´ .
With a look at Lemma 1 we estimate the remainder by
E[∣R(u)12] 6 16κ(u)-4(u4 + u2)2 E[∣F(Oi — O)(u)∣4 + ∣F(O) — Oi)(u)14]
. e2Tσmaxu u8(∆8 + ∆4∣∣tf∣∣42´ .
The Plancherel identity and these estimates yield together (44) via
E E[∣^(χ) — μ(χ) 12] dχ . U
-∞
2s +E(Tσ2U2)U5ε2 +E(2Tσm2axU2)U9ε4
+ E(Tσ2U2)U4ε2 +E(Tσm2axU2)2U9ε4
2s +E(σ2U2)U5ε2 +E(2Tσm2axU2)U9ε4.
7. Proof of the lower bounds
We follow the usual Bayes prior technique, see e.g. Korostelev and Tsybakov
(1993), and perturb a fixed Levy triplet T0 = (0,γ0,ν0) in the interior of
Gs (R, σmax) such that the perturbations remain in Gs(R, σmax).
More intriguing information
1. Effects of red light and loud noise on the rate at which monkeys sample the sensory environment2. The name is absent
3. SOME ISSUES CONCERNING SPECIFICATION AND INTERPRETATION OF OUTDOOR RECREATION DEMAND MODELS
4. CONSUMER PERCEPTION ON ALTERNATIVE POULTRY
5. AMINO ACIDS SEQUENCE ANALYSIS ON COLLAGEN
6. The name is absent
7. The name is absent
8. How we might be able to understand the brain
9. Innovation Trajectories in Honduras’ Coffee Value Chain. Public and Private Influence on the Use of New Knowledge and Technology among Coffee Growers
10. Transgression et Contestation Dans Ie conte diderotien. Pierre Hartmann Strasbourg