.png){kind=icon}
.png){kind=icon}
Return
GAUSSIAN-PROCESSMODELS OFPOPULATION DYNAMICS
.png){kind=icon}
PRE
AI
DOI：10.1615/JMachLearnModelComput.2025061979.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Surrogate models of population dynamics are crucial for ensemble-based computations such as dataassimilation, uncertainty quantification, and optimal design. One such surrogate is a data-drivenGaussian process (GP) that respects natural physical constraints. Our approach leverages outputtransformation and variational inference to ensure physical admissibility of the population variable.It also employs statistical learning and kernel discovery algorithms to identify GP structures thatachieve metric-based optimal fitting or prediction of the measured population data under differentenvironmental conditions. We examine the impact of different metrics on the resulting optimal ker-nels and propose an adaptive beam search method for efficient identification of such kernels. Theefficacy of the GP models is demonstrated through numerical experiments with a dataset describingthe evolution of a mosquito population. The proposed framework can be applied and extended toa wide range of biological systems, paving the way for uncertainty-informed decision making andreal-time data generation.
Keywords:
constrained Gaussian process
biological population dynamics
probabilistic forecasting
kernel structure learning
model validation metric
Journal
J
IF:
0
Papers:
8
Citations:
0