Corleno/Collaborative_Nonstationary_Multivariate_Gaussian_Process
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
We provide the main codes for our model in nmgp_dsvi.py, utils.py and pre_nmgp.py. For synthetic experiments, we provide the code for our model in "NMGP_SIM.ipynb" and the code for comparator models in "IGPR.ipynb" and "ICM.ipynb". The simulation codes are available in the folder "SIM_code" and the synthetic data in the paper are in data/simulation. Due to the size of supplementary we do not provide the real data but provide the codes. The data processing codes are in notebook/PM25.ipynb(HCP.ipynb, ECoG.ipynb); The codes for prediction task on the testing set are NMGP_PM25.py(NMGP_HCP.py, NMGP_ECoG_pred.py) and the code for posterior inference on the whole ECoG is NMGP_ECoG.py; The codes for comparators are in notebook/baseline_PM25.ipynb(baseline_HCP.ipynb,baseline_ECoG.ipynb).