Dissertation > Excellent graduate degree dissertation topics show

Expirical Likelihood for Partially Linear Models under Φ-Mixing Error

Author: HeXingXing
Tutor: WangChengMing
School: Guangxi Normal University
Course: Probability and Statistics
Keywords: Fixed Design φ- mixing error Partially Linear Models Empirical Likelihood Nonparametric likelihood ratio Weight function
CLC: O212.1
Type: Master's thesis
Year: 2005
Downloads: 77
Quote: 0
Read: Download Dissertation

Abstract


Empirical likelihood is Owen (1988) proposed under the sample in a completely non-parametric statistical inference method, which has similar characteristics Bootstrap sampling. This approach with classic or modern statistical methods has many prominent comparative advantages, such as: using empirical likelihood method to construct confidence intervals have domain retention, transformation invariance, confidence regions by the data to decide the shape and Bartlett Corrective and without structural axes statistics and many other advantages. Because of this, this method has aroused the interest of many statisticians, they will apply this approach to a variety of statistical models, such as linear models, generalized linear models, partially linear models, skewness sampling model, regression function, sub-bit estimates, M-functional, Kernel Density Estimation, biased sample, hate parameters, time series, conditional quantile and conditions density. But all of these are in the sample iid case be discussed, Kitamura, Kimchi (1997) The empirical likelihood applied to the weak dependency situation, Zhang warship, King fame, Wang Wei Xin (1999), respectively, in the m-dependent, α-mixing and φ-mixing case, the empirical likelihood studied, obtained similar results when independent and identically distributed. The partially linear model is Engle.et al (1986) in a study of climate conditions on the real problems affecting electricity demand raised. The statistics branch of the rise in recent years, both in practical application and theoretical research, which are subject to a number of scholars. Regression model with other issues, people interested in this subject of theoretical research focused on large sample nature, and since the early eighties has achieved fruitful results: the main research in a variety of assumptions β asymptotically efficient Estimated construction, β-weighted least squares estimation (?) n asymptotic normality, β estimated covariance function asymptotic properties, β and (?) n (g estimates) the optimal convergence speed strength β and g estimates nature. The past two years, China's scholars in the asymptotic statistical validity, M-estimation of asymptotic normality, parameter estimation component of the asymptotic distribution Berry-Essen bounds and iterated logarithm on other aspects of the study also received some fairly deep results. Shi.Jian, Lau.TaiShing etc. Empirical Likelihood this model has been discussed, but it is also based on a sample iid situation. This model is extended to fixed this design φ-mixing error situations and circumstances discussed in this regression coefficients empirical likelihood ratio estimation and confidence regions. Consider the following partial linear model y = x'β g (t) ε (2.1) where (x, t) ∈ R ~ p × R, t support set of bounded closed set, it may be set to [0,1] , g is defined in [0,1] on the unknown function, β ∈ R ~ p is an unknown regression coefficients to be estimated, y ∈ R is the dependent variable, ε ∈ R is unobservable random error. This article

Related Dissertations

  1. Based on empirical likelihood of p -order autoregressive model diagnostic statistics,O212.1
  2. The likelihood of missing data partially linear model heteroscedasticity test based on experience,O212.1
  3. Empirical Likelihood for Nonlinear Semi-parametric Regression Model with Missing Response Data,O212.1
  4. Empirical Likelihood for Nonlinear Semiparametric Models with m-dependent Errors,O212.1
  5. Empirical Likelihood for Nonparametric Models under Linear Process Errors,O212.7
  6. Empirical Likelihood Confidence Intervals for the Differences of Quantiles with Missing Data,O212.7
  7. Testing Hereoscedasticity in Varying-coefficient Models,O212.1
  8. Some Estimating Problems Based on Grouped Data,O212.1
  9. Statistical Analysis for Semiparametric Variable-Coefficient Partially Linear Errors-in-Variables with Missing Responses,O212.1
  10. Confidence Bands in the Semi-parametric Time Swries Regression,O211.67
  11. Based on the modified empirical likelihood variable selection method and its application,O211.67
  12. An Empirical Likelihood Goodness-of-Fit Test for Time Series,O211.61
  13. Application of EM Algorithm in Semiparametric Model with Missing Response,O212.1
  14. Varying-Coefficient Models Checking with Berkson Measurement Errors,O212.1
  15. Empirical Likelihood Inference for Ordinary Differential Equation Models with Error-prone State Variables,O212
  16. Variable Selection for High-dimensional Linear Modelsy,O212.7
  17. Properties of Estimators in Partially Linear Models under Restrictions,O212.1
  18. Empirical Likelihood Ratio Confidence Regions with a Nuisance Parameter for Dependent Samples,O212.1
  19. Empirical Likelihood Confidence Regions of Regression Coefficient in Quantile Regression Models under Φ-Mixing Samples,O212.1
  20. Empirical Likelihood Ratio Confidence Intervals for the Error Density in Linear Model,O212
  21. Missing data , linear regression statistical inference,O212.1

CLC: > Mathematical sciences and chemical > Mathematics > Probability Theory and Mathematical Statistics > Mathematical Statistics > General mathematical statistics
© 2012 www.DissertationTopic.Net  Mobile