Xingde Duan

Bio: Xingde Duan is an academic researcher from Guizhou University of Finance and Economics. The author has contributed to research in topics: Random effects model. The author has an hindex of 1, co-authored 2 publications receiving 1 citations. Previous affiliations of Xingde Duan include University of New Brunswick.

Assessing annual trends, monthly fluctuations and between-station relationship of sulphate deposition in the Turkey Lakes Watershed

Abstract: Water deposition of pollutants can be a good indicator of both air and water quality in a region of interest. In this paper, we study sulphate deposition change over time in a network of multiple monitoring stations in the Turkey Lakes Watershed in Sault Ste. Marie, Ontario, Canada. As there is generally substantial correlation among sulphate deposition observed over time and space, we incorporate temporally correlated multivariate random effects into Gamma regression models to account for the temporal dependence within each station and between-station dependence in space. We applied our new approach to analyse monthly average sulphate depositions between 1983 and 2003. We found the observed increase of sulphate deposition between 1994 and 2003 was not significant, that is, annual trends in sulphate deposition had stabilized since 1994. Our analysis also quantified increasing sulphate deposition from upstream to downstream and its monthly fluctuations from higher in winter to lower in summer. Understanding of these sulphate deposition trends is of great policy relevance to environmental conservation. Copyright © 2016 John Wiley & Sons, Ltd.

Assessing competition among species through simultaneously modeling marginal counts and respective proportions

Abstract: Evolution processes of multiple competitive and non-competitive species have traditionally been handled using different methods. In particular, evolution processes of multiple competitive species have usually been evaluated by the continuous and discrete proportions analysis; however, such evolution processes cannot be solely characterized by their relative proportions in practice. In this paper, we introduce a community based Poisson model with multivariate random effects to explicitly characterize marginal counts and respective proportions simultaneously. Furthermore, our method provides a unified approach to handle evolution processes of competitive and non-competitive species. In fact, the existence and strength of the competition among species can be assessed through our approach. Unlike those marginal modelling methods, our approach explicitly predicts random effects. Our model inference does not rely on distributional assumption of observed multivariate random effects, and thus is more robust than traditional approaches assuming parametric random effects.

A Joint Poisson State-Space Modelling Approach to Analysis of Binomial Series with Random Cluster Sizes.

Abstract: Serially correlation binomial data with random cluster sizes occur frequently in environmental and health studies. Such data series have traditionally been analyzed using binomial state-space or hidden Markov models without appropriately accounting for the randomness in the cluster sizes. To characterize correlation and extra-variation arising from the random cluster sizes properly, we introduce a joint Poisson state-space modelling approach to analysis of binomial series with random cluster sizes. This approach enables us to model the marginal counts and binomial proportions simultaneously. An optimal estimation of our model has been developed using the orthodox best linear unbiased predictors. This estimation method is computationally efficient and robust since it depends only on the first- and second- moment assumptions of unobserved random effects. Our proposed approach is illustrated with analysis of birth delivery data.