Let 0 [ a < b [ 2p and let D= def {e ih : h ¥ [a, b]}. We show that for generalized (non-negative) polynomials P of degree r and p > 0, we have where a 1 , a 2 , ..., a m ¥ D, c is an absolute constant (and, thus, it is independent of a, b, p, m, r, P, {a j }) and y is an explicitly determined cons
A large sieve density estimate near σ=1
✍ Scribed by P. X. Gallagher
- Publisher
- Springer-Verlag
- Year
- 1970
- Tongue
- English
- Weight
- 368 KB
- Volume
- 11
- Category
- Article
- ISSN
- 0020-9910
No coin nor oath required. For personal study only.
📜 SIMILAR VOLUMES
As alternatives to maximum likelihood estimation, conditional least squares (CLS) and maximum quasilikelihood estimation, with asymptotic properties, are investigated using a simpliÿed NEAR(1) model. The maximum quasilikelihood estimator seems to perform better than the CLS estimator in simulations.
The goal is to prove large deviations limit theorems for statistics, which are based on kernel density estimator and which are designed for symmetry testing. The formulas for the rate functions of the pointwise di erence and the uniform norm of the di erence are expressed in terms of the underlying