STELAR: Spatio-temporal Tensor Factorization with Latent Epidemiological Regularization
Authors: Nikos Kargas, Cheng Qian, Nicholas D. Sidiropoulos, Cao Xiao, Lucas M. Glass, Jimeng Sun4830-4837
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conduct experiments using both county- and state-level COVID-19 data and show that our model can identify interesting latent patterns of the epidemic. Finally, we evaluate the predictive ability of our method and show superior performance compared to the baselines, achieving up to 21% lower root mean square error and 25% lower mean absolute error for county-level prediction. |
| Researcher Affiliation | Collaboration | Nikos Kargas1,2,*, Cheng Qian2,*, Nicholas D. Sidiropoulos3, Cao Xiao2, Lucas M. Glass2,4, Jimeng Sun5 1 Dept. of ECE, University of Minnesota, 2 Analytics Center of Excellence, IQVIA, 3 Dept. of ECE, University of Virginia, 4 Dept. of Statistics, Temple University 5 Dept. of CS, University of Illinois Urbana-Champaign |
| Pseudocode | Yes | Algorithm 1 STELAR Method |
| Open Source Code | No | The paper does not provide an explicit statement or link for open-source code of the described methodology. It mentions the IQVIA patient claims dataset is "publicly accessible upon request" but not the code. |
| Open Datasets | Yes | We use US county-level data from the Johns Hopkins University (JHU) (Dong, Du, and Gardner 2020) and a large IQVIA patient claims dataset, which can be publicly accessible upon request. |
| Dataset Splits | Yes | Initially we use 85 days for training and validation and use the remaining Lo = 10 as the test set. In the second experiment we use 80 days for training and validation and Lo = 15 days for test. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers needed to replicate the experiment. |
| Experiment Setup | Yes | Input: Tensor X, rank K, max. outer iterations itersouter, max. inner iterations itersinner, gradient steps itersgrad, prediction window Lo. We train our model on county-level data using K = 30. |