Table 5.
Included studies covering spatio-temporal monitoring of disease spread with hybrid models and their key aspects.
| Study | Key aspects |
|---|---|
| Dandekar and Barbastathis (56) | Analyze NPIs in different countries to find effective reproduction number |
| Menda et al. (57) | Estimate dynamic transmission number with NN, allowing for multi-peaks |
| Silva et al. (58) | Build society with ABM and simulate different NPI scenarios |
| Capobianco et al. (59) | Combine ABM and SEIR with Markov model and RL for NPI planning |
| Wang et al. (60) | Combine spatial and temporal models |
| Watson et al. (61) | Predict deaths by relation between cases and population characteristics |
| Fritz et al. (62) | Use a GNN to include local mobility and connectedness data from Meta |
| Hadley et al. (63) | Modify transmission and hospitalization rates fitted to agent's characteristics |
NPI, non-pharmaceutical intervention; NN, neural network; ABM, agent-based modeling; SEIR, susceptible-exposed-infected-recovered; RL, reinforcement learning; GNN, graph neural network.