Description
1- Machine learning methods in reionization modeling and data analysis/inference. How can AI/ML tools (e.g. gaussian processes, convolutional neural networks, generative adversarial networks, graphs, symbolic regression) be used to improve theoretical modeling and data analysis/inference?
The next three topics are meant to connect theoretical modeling to observations by answering questions like: Have we extracted all the available information from existing observations to constrain models? Which (ideal) observations would we like to have? Are our models appropriate to correctly interpret observations?
2- Modeling high-z galaxies. Example topics include: Are the modeling techniques appropriate for interpreting the observations correctly? Do different modeling techniques lead to different inferences? Are there important missing ingredients in existing models which may bias inferences? Are there ambiguities in comparing measured and modeled quantities? Which observations are most informative to modelers? Are the models predictive? Can current simulations model the escape fraction?
3- Modeling QSO absorption spectra. Example topics include: Are the modeling techniques appropriate for interpreting the observations correctly? Do different modeling techniques lead to different inferences? Are there important missing ingredients in existing models which may bias inferences? Are there ambiguities in comparing measured and modeled quantities? Which observations are most informative to modelers? Are the models predictive?
4- Modeling line intensity mapping and the CMB. Are the modeling techniques appropriate for interpreting the observations correctly? Do different modeling techniques lead to different inferences? Are there important missing ingredients in existing models which may bias inferences? Are there ambiguities in comparing measured and modeled quantities? Which observations are most informative to modelers? Are the models predictive?
