AAAI22

• PDF
• Author Response during the rebuttal
• Appendix

Setup

1. Install nsgp-torch package

pip install git+https://github.com/patel-zeel/nsgp-torch

2. Install other dependencies

pip install -r requirements.txt

3. Run the experiments from individual folders.

Main approach configuration

Encoding

A - ARD enabled

A_bar - ARD disabled

N - Non-stationary kernel

N_bar - Stationary kernel

C - Using categorical kernel for categorical features without one-hot-encoding

C_bar - Using RBF/Matern kernel for categorical features with one-hot-encoding

L - Using Local periodic kernel for time feature

L_bar - Using RBF/Matern kernel for time feature

Example

AN_barCL_bar - GP with ARD enabled stationary kernel with categorical kernel for categorical features and RBF/Matern kernel for time feature

Folder-wise description

Folder	Description
data	data for each baseline and main approach
preprocessing	preprocessing pipeline applied to data
stat_gp_cat	Stationary GP with categorical kernel (C fixed, L variable)
stat_gp_no_cat	Stationary GP without categorical kernel (C_bar fixed, L variable)
nonstat_gp_cat	Non-stationary GP with categorical kernel (C fixed, L variable)

Baselines

Baseline implementation of paper "A Neural Attention Model for Urban Air Quality Inference: Learning the Weights of Monitoring Stations" (ADAIN) is available in this file.