This paper studies how to detect change
points in IV models. LM method is proposed based on weighted
residual partial sum. The asymptotic properties under null or
alternative hypothesis are proved. Under some non-orthogonality
condition, the test has non-trivial power when \delta=1/2 and will
be consistent when 0<\delta<1/2.
In this paper, we firstly derived the expressions
of the well-known ordinary least square estimator (OLSE), the
ordinary mixed estimator (OME) introduced by Theil and Golberger
(1961) and the stochastic restricted Liu estimator (SRLE) proposed
by Yang and Xu (2007) under misspecification due to inclusion of
some superfluous variables. Then, performances of these estimators
under misspecification are examined. In particular, necessary and
sufficient conditions for the superiority of the SRLE over the OLSE
and OME with respect to the mean squared error matrix (MSEM)
criterion are derived. Furthermore, superiority of the corresponding
predictors of these estimators are also investigated.
Hypertensive disorder complicating pregnancy
(HDCP) is a particular disease in the pregnancy. With risk factors
appear, such as the fast pace of life, mental stress, the increasing
of the elderly primipara, the number of women with HDCP tends to
have increased gradually. At present, there are a lot researches
about the risk factors of HDCP. In this article, we make variable
selection for the risk factors based the double logistic regression
models and give a method about forecasting at last.
A likelihood approach, together with
a EM-type algorithm,to jointly estimate the regression coefficient
as well as the marginal distribution of the covariant in regression
model with an interval-censored data covariant is developed. Under
certain conditions the procedures are convergent, and the resulting
estimators are asymptotically normal.
Based on the definition of supporting degree,
we put forward the definitions of average index-supporting degree
and average dispersion, and we build an optimal combination
forecasting model based on exponential supporting degree. We also
give its equivalent average dispersion optimal combination
forecasting model. We propose the conception of superior combination
forecasting, we obtain the results, including the sufficient
conditions for the existing non-inferior and superior combination
forecasting, redundant forecasting and the redundant information
determination. Finally, we give the example to show that the model
is effective.
Under simple hypothesis, we show that the delta-corrected Cram\'er-von Mises test is non-unbiased for any
given level of significance $\alpha\in(0,1)$ and any sample size n.
The pricing of the derivatives associated
with counterparty default risk is considered. Based on Merton's
structured credit risk model, an explicit pricing formula of
vulnerable options was derived when the underlying asset price and
corporate value is assumed to follow a jump-diffusion process. A
model of vulnerable option pricing is developed when the underlying
asset price and corporate value is assumed to follow a
jump-diffusion process, then the pricing of vulnerable option is
discussed when the corporate liabilities are fixed and random were
derived respectively.
In this paper, the almost sure convergence
and complete convergence for $\wt{\varphi}$-mixing random variables
are established. The results obtained not only extend and generalize
the classical Khintchine-Kolmogorov Convergence Theorem, the Three
Series Theorem for independent random variables to the case of
$\wt{\varphi}$-mixing random variables, but also improve the
relevant results without necessarily adding any extra any
conditions.
In classical credibility theory,
the risks in a portfolio are assumed to be mutually independent and
the premiums are derived under squared loss functions. In this
paper, we develop the credibility theory under balanced loss
functions with a special dependence structure among the individual
risks: induced by common effect (Wen et al., 2009). To be specific,
credibility premiums under balanced loss functions with common
effects are derived for B\"{u}hlmann and B\"{u}hlmann-Straub
credibility models.
In this paper, the Fisher scoring method
is applied to get M-estimator (robust estimator) in the mixed
effects linear model for longitudinal data. And its asymptotic
property is given later. Then the local influence analysis of
variance of the error is studied based on M-estimation. At last the
method is illustrated by the grape sugar data example.