A nonparametric procedure is proposed to
detect multiple changes in the mean of independent random series and
the asymptotic distribution is derived. Simultaneously, the
estimators for the locations of the change points are obtained.
Moreover, the performance of the test is studied by Monte Carlo
simulation, which demonstrates that the proposed test has high
powers and good sizes for heavy-tailed innovations. Finally, the
feasibility of the proposed test is illustrated by the application
on the LBC data of stock prices.
In this paper, the diagnostic measures for
censored linear models are studied based on the empirical likelihood
method. First, the diagnostic measures for linear models are
studied; Then, the censored linear models are converted to linear
models, and the diagnostic measures for converted models are
studied; Last, simulation studies and real data analysis are given
to illustrate the validity of statistical diagnostic measures.
In this paper, the Fisher scoring method
is applied to get M-estimator (robust estimator) in the mixed
effects linear model for longitudinal data. The score tests for
correlation coefficients in the model with uniform correlation
covariance structure based on M-estimator are also studied. Then the
properties of test statistics are investigated through Monte Carlo
simulations. At last, the methods and properties are illustrated by
the grape sugar data example.
This paper focuses on ruin probability for
Cox model with variable premium rate and constant investment return
when the claims have heavy tailed distribution. By considering the
"skeleton process'' of Cox risk model, a recursive equation for
finite time ruin probabilities are derived in terms of "renewal
techniques'' and asymptotic estimation for finite time ruin
probabilities and ultimate ruin probability are obtained by
inductive method.
We generalize the well-known Dobrushin
coefficient in total variation to weighted total variation
, which gives a criterion for the geometric ergodicity of
discrete-time Markov chains.
In this paper, we study absolute ruin
problems for the Sparre Andersen risk process with generalized
Erlang()-distributed inter-claim times, investment and debit
interest. We first give a system of integro-differential equations
with certain boundary conditions satisfied by the expected
discounted penalty function at absolute ruin. Second, we obtain a
defective renewal equation under some special cases, then based on
the defective renewal equation we derive two asymptotic results for
the expected discounted penalty function when the initial surplus
tends to infinity for the light-tailed claims and heavy-tailed
claims, respectively. Finally, we investigate some explicit
solutions and numerical results for generalized Erlang(2)
inter-claim times and exponential claims.
In this paper, we compare two modified
Gaussian pseudolikelihood criteria (GPCs) with existing Gaussian
pseudolikelihood criterion and empirical likelihood based criteria
to choose the working correlation matrix in generalized estimating
equations approach. Rich simulation studies are conducted to
investigate the performance of these criteria under a range of model
settings. The results show that the modified criteria outperform the
original GPC and empirical likelihood based criteria in most cases
in terms of selection accuracy. Empirical likelihood based criteria
perform better to identify exchangeable structure in data with
binary response. In the end, these criteria are applied to epilepsy
seizure and Madras longitudinal schizophrenia study clinical data
sets analysis.
This paper considers the valuation of
guaranteed minimum death benefit in variable annuities under a
regime-switching model. More specifically, the risk-free interest
rate, the appreciation rate and the volatility of the reference
investment fund are modulated by a continuous-time, finite-state,
observable Markov chain. A regime-switching Esscher transform is
adopted to select an equivalent martingale measure in the incomplete
financial market. Inverse Fourier transform is used to derive an
analytical pricing formula for the embedded option in variable
annuity with guaranteed minimum death benefit. To calculate the fair
guarantee charge, fast Fourier transform approach is applied.
Numerical examples are provided to illustrate the practical
implementation and the relationship between the fair guarantee
charges and other parameters.
In this paper, we consider the optimal
joint dividend and capital injection strategy with proportional and
fixed costs. It supposes that capitals can be injected whenever they
are profitable, but dividends can only be paid at the arrival times
of a Poisson process with intensity . Our objective is to
determine an optimal strategy of maximizing the expected cumulative
discounted dividends minus the expected discounted costs of capital
injections before bankruptcy. By solving some impulse problems, we
get the closed-form solutions depending on the parameters of model.
Some known results in Lokka and Zervos (2008) can be viewed as
limiting cases when .