The HEP cross-frontier machine learning group (HXML) is a collaborative research initiative at SLAC national accelerator laboratory, led by Michael Kagan, Kazu Terao and Phil Marshall. One of our activities is this reading group, the Breakfast Club.
We meet by video every two weeks (on Tuesdays, at 0900 PT, hence the group/repo name) to discuss a paper that we have all agreed to read: the goal is to learn new machine learning methods and approaches together, and build up a shared, annotated bibliography file, hxml.bib. Each paper has an assigned "lead" reader, who has agreed to lead the discussion of the paper. We plan to spend a few months on each topic, and then move onto a new one. You can see what we are currently reading in the tables below; suggestions for new papers, and new topics, are made in the issues.
The reading group meetings are only open to HXML group members, but our bibliography file is available for anyone to use under the creative commons CC0 license. Feel free to send us questions and comments via the issues!
Winter 2020
| Paper (link to relevant issue with slides and paper) | Date | Leader |
|---|---|---|
| Hierarchical Implicit Models and Likelihood-Free Variational Inference | 1/28/2020 | @swagnercarena |
| Variational Inference with Normalizing Flows | 3/10/2020 | @jiwoncpark |
| Automatic Posterior Transformation for Likelihood-free Inference | 3/24/2020 | @swagnercarena |
| On Contrastive Learning for Likelihood-free Inference | 6/03/2020 | @jiwoncpark |
| Empirical Bayes for Likelihood-free Inference (his own application) | 6/24/2020 | @MaximeVandegar |
| Normalizing Flows for Probabilistic Modeling and Inference | 7/08/2020, 07/29/2020 | @jiwoncpark, @joshualin24 |
| Bayesian Deep Learning and a Probabilistic Perspective of Generalization | 8/12/2020 | @swagnercarena |
| A general method for debiasing a Monte Carlo estimator | 9/09/2020 | @MaximeVandegar |
| Importance Weighted Hierarchical Variational Inference | 9/30/2020 | @jiwoncpark |