Beyond Intuition, a Framework for Applying GPs to Real-World Data
This work addresses the challenge of applying GPs to complex datasets for practitioners, though it is incremental as it formalizes existing expert knowledge.
The authors tackled the problem of deploying Gaussian Processes (GPs) in real-world scenarios by proposing a framework to assess suitability and set up robust models, resulting in more accurate predictions in a glacier elevation change case study.
Gaussian Processes (GPs) offer an attractive method for regression over small, structured and correlated datasets. However, their deployment is hindered by computational costs and limited guidelines on how to apply GPs beyond simple low-dimensional datasets. We propose a framework to identify the suitability of GPs to a given problem and how to set up a robust and well-specified GP model. The guidelines formalise the decisions of experienced GP practitioners, with an emphasis on kernel design and options for computational scalability. The framework is then applied to a case study of glacier elevation change yielding more accurate results at test time.