scholarly journals Parametric inference for discrete observations of diffusion processes with mixed effects

2018 ◽  
Vol 128 (6) ◽  
pp. 1929-1957 ◽  
Author(s):  
Maud Delattre ◽  
Valentine Genon-Catalot ◽  
Catherine Larédo
Keyword(s):  
Diffusion Processes ◽  
Mixed Effects ◽  
Parametric Inference ◽  
Discrete Observations

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

2015 ◽  
Vol 08 (05) ◽  
pp. 1550060 ◽  
Author(s):  
Petras Rupšys
Keyword(s):  
Fixed Effects ◽  
Diffusion Processes ◽  
Random Parameter ◽  
Mixed Effects ◽  
Validation Dataset ◽  
Natural Systems ◽  
Breast Height ◽  
Deterministic Function ◽  
Pine Trees ◽  
Soil Fertility Index

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, c h , crown width, c w , mean breast height diameter, d0, mean height, h0, 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 height–diameter 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 0 cm. 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.

Bayes factors for discrete observations from diffusion processes

Biometrika ◽  
1994 ◽  
Vol 81 (1) ◽  
pp. 11-26 ◽  
Author(s):  
N. G. POLSON ◽  
G. O. ROBERTS
Keyword(s):  
Diffusion Processes ◽  
Bayes Factors ◽  
Discrete Observations
Sign in / Sign up

Export Citation Format

Share Document