APPROXIMATION RATES OF ERROR DISTRIBUTION OF DOUBLE KERNEL ESTIMATES OF CONDITIONAL DENSITY

In this paper, the normal approximation rate and the random weighting approximation rate of error distribution of the kernel estimator of conditional density function f(y|x) are studied. The results may be used to construct the confidence interval of f(y|x) .

Approximating Conditional Density Functions Using Dimension Reduction

We propose to approximate the conditional density function of a random variable Y given a dependent random d-vector X by that of Y given θ^τX, where the unit vector θ is selected such that the average Kullback-Leibler discrepancy distance between the two conditional density functions obtains the minimum. Our approach is nonparametric as far as the estimation of the conditional density functions is concerned. We have shown that this nonparametric estimator is asymptotically adaptive to the unknown index θ in the sense that the first order asymptotic mean squared error of the estimator is the same as that when θ was known. The proposed method is illustrated using both simulated and real-data examples.

Asymptotic properties of a nonparametric conditional density estimator in the local linear estimation for functional data via a functional single-index model

This paper deals with the conditional density estimator of a real response variable given a functional random variable(i.e.,takes values in an infinite-dimensional space).Specifically,we focus on the functional index model,and this approach represents a good compromise between nonparametric and parametric models.Then we give under general conditions and when the variables are independent,the quadratic error and asymptotic normality of estimator by local linear method,based on the single-index structure.Finally,wecomplete these theoretical advances by some simulation studies showing both the practical result of the local linear method and the good behaviour for finite sample sizes of the estimator and of the Monte Carlo methods to create functional pseudo-confidence area.

RE-WEIGHTED NADARAYA-WATSON ESTIMATION OF CONDITIONAL DENSITY

In order to avoid the discussion of equation （1.1）, this paper employs the proof method of Liang （2012） to consider the re-weighted Nadaraya-Watson estimation of conditional density. The established results generalize those of De Gooijer and Zerom （2003）. In addition, this paper improves the bandwidth condition of Liang （2012）.

Strong Convergence Rates of Double Kernel Estimates of Conditional Desity Under Stationary Sequences

In the paper,we study the strong convergence rates of double kernel estimates of conditional density under stationary sequences.

Nonparametric inferences for kurtosis and conditional kurtosis

Under the assumption of strictly stationary process, this paper proposes a nonparametric model to test the kurtosis and conditional kurtosis for risk time series. We apply this method to the daily returns of S＆P500 index and the Shanghai Composite Index, and simulate GARCH data for verifying the efficiency of the presented model. Our results indicate that the risk series distribution is heavily tailed, but the historical information can make its future distribution light-tailed. However the far future distribution＇s tails are little affected by the historical data.

NEW METHOD TO OBTAIN THE POWER SPECTRA OF HIDDEN VARIABLES AND ITS APPLICATION TO OCEAN DATA

A novel method is proposed to obtain the power spectra of hidden variables in a chaotic time series. By embedding the data in phase space , and recording the conditional probability density of points that the trajectory encounters as it evolves in the reconstructed phase space, it is possible to recover the power spectra of hidden variables in chaotic time series through a spectral analysis over the conditional probability density time series. The method is robust in the application to Lorenz system, 4 dimension Rssler system and rigid body motion by linear feedback system (LFRBM). Applying the method to the time series of sea surface temperature (SST) of the South China Sea, we obtained the power spectra of the wind speed (WS) from SST data. Furthermore, the results showed that there exists an important nonlinear interaction between the SST and the WS.

DIAGNOSIS AND IDENTIFICATION OF DYNAMIC CORRELATION FACTORS BETWEEN WEST-PACIFIC SUBTROPICAL HIGH AND EAST ASIA MONSOON SYSTEM INDEXES

Based on the daily reanalysis data of NCEP / NCAR and by using the method of phase space reconstruction, the point conditional probability density of the subtropical high ridge index are determined and then used, together with their power spectra, to seek the correlation between them and individual monsoon-affecting factors and their power spectra. Through diagnosis, six indexes are discovered that have the most important effects on the subtropical high index. The results of the diagnosis indicate that the technique can identify the factors which are dynamically correlated. It can offer the basis in determining and choosing dynamic conceptual factors.

Density Results for Subspace Multiwindow Gabor Systems in the Rational Case

Let S be a periodic set in R and L2（S） be a subspace of L2（R）. This paper investigates the density problem for multiwindow Gabor systems in L2（S） for the case that the product of time- frequency shift parameters is a rational number. We derive the density conditions for a multiwindow Gabor system to be complete （a frame） in L2（S）. Under such conditions, we construct a multiwindow tight Gabor frame for L2 （S） with window functions being characteristic functions. We also provide a characterization of a multiwindow Gabor frame to be a Riesz basis for L2（S）, and obtain the density condition for a multiwindow Gabor Riesz basis for L2 （S）.

Tree-Indexed Markov Chains in Random Environment and Some of Their Strong Limit Properties

In this paper,the authors first introduce the tree-indexed Markov chains in random environment,which takes values on a general state space.Then,they prove the existence of this stochastic process,and develop a class of its equivalent forms.Based on this property,some strong limit theorems including conditional entropy density are studied for the tree-indexed Markov chains in random environment.

A new inverse method and application to ocean data

A new method is proposed to inverse normalization data of hidden variables in a dynamical system by embedding a time series in multidimensional spaces and applying a normalization analysis to the conditional probability density of points in the reconstructed phase spaces. The method is robust in the application to Lorenz system and 4-dimensional R?ssler system by testing quantitatively and qualitatively the correlation coefficient between inverse data and original data in time domain and in frequency domain, respectively. By applying the method to analyzing the South China Sea data, the normalization data of wind speed is extracted from the sea surface temperatu-re time series.

Generalized fixed-effects and mixed-effects parameters height-diameter models with diffusion processes

Statistical models using stochastic differential equations （SDEs） to describe dynamically evolving natural systems are appearing in the scientific literature with some regularity in recent years. In this study, the SDE mixed-effects parameter models based on a Vasicek non-homogeneous diffusion process are formulated. The breast height diameter- dependent drift function additionally depends on deterministic function that describes the dynamic of certain exogenous stand variables （crown height, eh, crown width Cw, mean breast height diameter, do, mean height, ho, age, A, soil fertility index SFI, stocking level, S） versus breast height diameter. The mixed-effects parameters SDE models included a random parameter that affected the models asymptote. The parameter estimators are evaluated by maximum likelihood procedure. The objective of the research was to develop a generalized mixed-effects parameters SDE heightdiameter models and to illustrate issues using dataset of Scots pine trees （Pinus sylvestris L.） in Lithuania with the breast height diameter outside the bark larger than 0cm. The parameters of all used models were estimated using an estimation dataset and were evaluated using a validation dataset. The new developed height diameter models are an improvement over exogenous stand variables, in that it can be calibrated to a new stand with observed height-diameter pairs, thus improving height prediction.