In this paper, we analyze the time series of site coordinates of 27 continuously monitoring GPS sites covered by the Crustal Movement Observation Network of China over the whole country. The data are obtained in the p...In this paper, we analyze the time series of site coordinates of 27 continuously monitoring GPS sites covered by the Crustal Movement Observation Network of China over the whole country. The data are obtained in the period from the beginning of the observation to the November of 2005. On the basis of data processing, we analyze the power spectrum density of coordinate component noise at each site and calculate the spectral indexes manifesting the noise property of each component. The spectral indexes indicate that for most sites, the noise of time series of each coordinate component can be addressed by the model of white noise ＋ flicker noise; and for a small amount of sites, it can be described by the model of white noise ＋ flicker noise ＋ random walk noise. We also quantitatively estimate each noise component in the model by using the criterion of maximum likelihood estimation. The result shows that the white noise in the time series of GPS site coordinates does not constitute the main part of noise. Therefore, the error estimation of site movement parameters is usually too small, or too optimistic if we consider the white noise only. Correspondingly, if this factor is not fully considered in explaining these movement parameters, it might mislead the readers.展开更多
This paper proposes a simple constant-stress accel- erated life test （ALT） model from Burr type XII distribution when the data are Type-I progressively hybrid censored. The maximum likelihood estimation （MLE） of t...This paper proposes a simple constant-stress accel- erated life test （ALT） model from Burr type XII distribution when the data are Type-I progressively hybrid censored. The maximum likelihood estimation （MLE） of the parameters is obtained through the numerical method for solving the likelihood equations. Approxi- mate confidence interval （CI）, based on normal approximation to the asymptotic distribution of MLE and percentile bootstrap Cl is derived. Finally, a numerical example is introduced and then a Monte Carlo simulation study is carried out to illustrate the pro- posed method.展开更多
The estimation of generalized exponential distribution based on progressive censoring with binomial removals is presented, where the number of units removed at each failure time follows a binomial distribution. Maximu...The estimation of generalized exponential distribution based on progressive censoring with binomial removals is presented, where the number of units removed at each failure time follows a binomial distribution. Maximum likelihood estimators of the parameters and their confidence intervals are derived. The expected time required to complete the life test under this censoring scheme is investigated. Finally, the numerical examples are given to illustrate some theoretical results by means of Monte-Carlo simulation.展开更多
One-bit feedback systems generate binary data as their output and the system performance is usually measured by the success rate with a fixed parameter combination. Traditional methods need many executions for paramet...One-bit feedback systems generate binary data as their output and the system performance is usually measured by the success rate with a fixed parameter combination. Traditional methods need many executions for parameter optimization. Hence, it is impractical to utilize these methods in Expensive One-Bit Feedback Systems （EOBFSs）, where a single system execution is costly in terms of time or money. In this paper, we propose a novel algorithm, named Iterative Regression and Optimization （IRO）, for parameter optimization and its corresponding scheme based on the Maximum Likelihood Estimation （MLE） method and Particle Swarm Optimization （PSO） method, named MLEPSO-IRO, for parameter optimization in EOBFSs. The IRO algorithm is an iterative algorithm, with each iteration comprising two parts： regression and optimization. Considering the structure of IRO and the Bernoulli distribution property of the output of EOBFSs, MLE and a modified PSO are selected to implement the regression and optimization sections, respectively, in MLEPSO-IRO. We also provide a theoretical analysis for the convergence of MLEPSO-IRO and provide numerical experiments on hypothesized EOBFSs and one real EOBFS in comparison to traditional methods. The results indicate that MLEPSO-IRO can provide a much better result with only a small amount of system executions.展开更多
基金Special project of China Earthquake Administration"Study on the Integrated Observation of Vertical Crustal Move-ment and Deformation of South-North Seismic Zone on the Chinese Mainland".
文摘In this paper, we analyze the time series of site coordinates of 27 continuously monitoring GPS sites covered by the Crustal Movement Observation Network of China over the whole country. The data are obtained in the period from the beginning of the observation to the November of 2005. On the basis of data processing, we analyze the power spectrum density of coordinate component noise at each site and calculate the spectral indexes manifesting the noise property of each component. The spectral indexes indicate that for most sites, the noise of time series of each coordinate component can be addressed by the model of white noise ＋ flicker noise; and for a small amount of sites, it can be described by the model of white noise ＋ flicker noise ＋ random walk noise. We also quantitatively estimate each noise component in the model by using the criterion of maximum likelihood estimation. The result shows that the white noise in the time series of GPS site coordinates does not constitute the main part of noise. Therefore, the error estimation of site movement parameters is usually too small, or too optimistic if we consider the white noise only. Correspondingly, if this factor is not fully considered in explaining these movement parameters, it might mislead the readers.
基金supported by the National Natural Science Foundation of China(7117116470471057)
文摘This paper proposes a simple constant-stress accel- erated life test （ALT） model from Burr type XII distribution when the data are Type-I progressively hybrid censored. The maximum likelihood estimation （MLE） of the parameters is obtained through the numerical method for solving the likelihood equations. Approxi- mate confidence interval （CI）, based on normal approximation to the asymptotic distribution of MLE and percentile bootstrap Cl is derived. Finally, a numerical example is introduced and then a Monte Carlo simulation study is carried out to illustrate the pro- posed method.
基金supported by the National Natural Science Foundation of China(70471057)
文摘The estimation of generalized exponential distribution based on progressive censoring with binomial removals is presented, where the number of units removed at each failure time follows a binomial distribution. Maximum likelihood estimators of the parameters and their confidence intervals are derived. The expected time required to complete the life test under this censoring scheme is investigated. Finally, the numerical examples are given to illustrate some theoretical results by means of Monte-Carlo simulation.
文摘One-bit feedback systems generate binary data as their output and the system performance is usually measured by the success rate with a fixed parameter combination. Traditional methods need many executions for parameter optimization. Hence, it is impractical to utilize these methods in Expensive One-Bit Feedback Systems （EOBFSs）, where a single system execution is costly in terms of time or money. In this paper, we propose a novel algorithm, named Iterative Regression and Optimization （IRO）, for parameter optimization and its corresponding scheme based on the Maximum Likelihood Estimation （MLE） method and Particle Swarm Optimization （PSO） method, named MLEPSO-IRO, for parameter optimization in EOBFSs. The IRO algorithm is an iterative algorithm, with each iteration comprising two parts： regression and optimization. Considering the structure of IRO and the Bernoulli distribution property of the output of EOBFSs, MLE and a modified PSO are selected to implement the regression and optimization sections, respectively, in MLEPSO-IRO. We also provide a theoretical analysis for the convergence of MLEPSO-IRO and provide numerical experiments on hypothesized EOBFSs and one real EOBFS in comparison to traditional methods. The results indicate that MLEPSO-IRO can provide a much better result with only a small amount of system executions.
