phenoCDM: Continuous Development Models for Incremental Time-Series Analysis
- 1. Har
- 2. Duke University
Description
Using the Bayesian state-space approach, we developed a continuous development model to quantify dynamic incremental changes in the response variable. While the model was originally developed for daily changes in forest green-up, the model can be used to predict any similar process. The CDM can capture both timing and rate of nonlinear processes. Unlike statics methods, which aggregate variations into a single metric, our dynamic model tracks the changing impacts over time. The CDM accommodates nonlinear responses to variation in predictors, which changes throughout development.
Files
bnasr/phenoCDM-v0.1.1.zip
Files
(364.8 kB)
Name | Size | Download all |
---|---|---|
md5:796c0a0cdd19baaeaf29d78a996d0127
|
364.8 kB | Preview Download |
Additional details
Related works
- Is supplement to
- https://github.com/bnasr/phenoCDM/tree/v0.1.1 (URL)