Abstract
A single-index varying coefficients regression model is considered, from which robust confidence regions/intervals of the regression parameters are derived. Considering the rank-based estimating equations, empirical likelihood objective functions of the index and functional coefficients objective functions are defined and their asymptotic properties are established under mild regularity conditions. The performance of the proposed approach is demonstrated via extensive Monte Carlo simulation experiments. The simulation results are compared with those obtained from a normal approximation alternative. Also the proposed method is compared with the least squares and least absolute deviations alternatives. Finally, a real data example is given to illustrate the method.
2000 AMS Subject Classifications:
Disclosure statement
No potential conflict of interest was reported by the author(s).
Appendix
The following lemma is necessary in the proof of Theorem 3.2
Lemma 4.1
Set , where
is the cumulative distribution function of
. Under
, we have
Proof.
Following [Citation32], it is not hard to show that
By continuity of φ together with the generalized continuous mapping theorem [Citation33], we have
On the other hand,
Note that by definition,
and by assumption
,
. Then, by Cauchy–Schwarz inequality, we have
By
and
,
. So, by the Strong Law of Large Numbers,
Moreover, as
,
Thus,
Proof of Theorem 3.1
By definition, satisfies
Now, by assumption
and
, one can obtain in a straightforward manner that
. Also,
. Following the same argument as in [Citation34], we obtain that
, and by Theorem 2.3 of [Citation23],
. From this, a Taylor expansion to the right hand side of Equation (Equation8
(8)
(8) ) gives
where
. On the other hand, note that solving Equation (Equation8
(8)
(8) ) is equivalent to solving
which in turn can be rewritten as
(17)
(17) This implies that
from which, it follows that
since
. Also, premultiplying Equation (Equation17
(17)
(17) ) by
, yields
which in turn results in
since
. Hence,
From this, note that by Theorem 2.3 of [Citation23], we have
Thus
Hence, noting that
, where
is a
identity matrix,
and the proof is complete.
Proof of Theorem 3.2
For fixed and from Equation (Equation13
(13)
(13) ),
satisfies
(18)
(18) Now, by Assumptions
and
, one can obtain in a straightforward manner that
. Also,
. Following the same argument as in the proof of Theorem 3.1, we have
, and by Lemma 3.1,
. Once again, a direct application of the Taylor expansion up to order 2 to the right-hand side of Equation (Equation14
(14)
(14) ) gives
where
. Following similar argument as in the proof of Theorem 3.1, solving Equation (Equation18
(18)
(18) ) is equivalent to solving
which in turn can be rewritten as
(19)
(19) Solving this equation for
gives
from which, it follows that
as
. Moreover, premultiplying Equation (Equation19
(19)
(19) ) by
, yields
which in turn results in
since
. Thus
Noting that
is symmetric, we have
To this end, from Remark 3.2,
, where
is a
identity matrix. Thus
, which completes the proof.