Bayesian Optimization through Gaussian Cox Process Models for Spatio-temporal Data

Authors: Yongsheng Mei, Mahdi Imani, Tian Lan

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive evaluations on various synthetic and real-world datasets demonstrate significant improvement over state-of-the-art inference solutions for Gaussian Cox processes, as well as effective BO with a wide range of acquisition functions designed through the underlying Gaussian Cox process model.
Researcher Affiliation Academia Yongsheng Mei1, Mahdi Imani2, Tian Lan1 1The George Washington University 2Northeastern Univeristy {ysmei, tlan}@gwu.edu, m.imani@northeastern.edu
Pseudocode Yes Algorithm 1 BO on the Gaussian Cox Process
Open Source Code Yes The code has been made available on Git Hub via https://github.com/ysmei97/gaussian cox bo.
Open Datasets Yes Extensive evaluations are conducted using both synthetic functions in the literature (Adams et al., 2009) and real-world spatio-temporal datasets, including DC crime incidents (DC.gov, 2022), 2D neuronal data (Sargolini et al., 2006), and taxi data of Porto (O Connell et al., 2015).
Dataset Splits No The paper discusses observed regions and initial observations, and iterates on adding new samples, but does not specify fixed training, validation, and test dataset splits with percentages or counts.
Hardware Specification Yes We conducted our experiments on the Ubuntu 20.04 system, with Intel(R) Core(TM) i7-6700 4-core CPU (3.4 GHz) and 16.0 GB RAM.
Software Dependencies Yes The algorithm is implemented in Python 3.8, using main Python libraries Num Py 1.22.3 and Pandas 2.0.3.
Experiment Setup Yes In the experiment, we initialize the region centers t = (25, 60) in the time domain of [0, 100] with a region radius of 2, i.e., region size of 4. Observations in selected regions are highlighted in red vertical bars. We use the UCB acquisition function for identifying intensity peaks. To expedite the BO process, we set the acquisition values of explored regions as zero to prevent the algorithm from becoming trapped within the same region.