I am currently spending the week in Leiden (Netherlands), attending to a very interesting workshop that brought together computer scientists, astronomers, astrophysicists, and particle physicists to discuss how to apply the most cutting-edge tools in machine learning to improve our chances of discovering dark matter, the unknown non-luminous substance that vastly overweights luminous matter in the Universe.
The venue is the Lorentz center of Leiden University, a wonderful place for workshops, where you are pampered in many ways, from having your private office, with digital key and all, to having a coffee room with armchairs and all sort of treats permanently available. Also, the weather this week couldn't have been better, as the hurricane-class 140km/h winds outside discouraged anybody from sneaking out of the workshop.
Talks were very inspiring and insightful, and the afternoons allowed people to get together and brainstorm about common projects. I learned a few things, but more importantly I made a few acquaintances that may open the way to future collaboration. As for the organizers, we already are participating in a common endeavour - a new ITN proposal we just submitted. So let's cross fingers for that.
The discussions and the talks were informal and sparkling, leaving room for lively discussions and improvisation. An example of what I mean is given in the picture below. Martin White is talking about ML ideas for DM model exploration, and all at once he picks up a trash bin and lays it upside down on the table in the front row.
Can you guess what he was saying when he did that ? Here are four possible captions, suggested after the fact by attendees.
1 - "Believe it or not, I have just trapped 53 WIMPS inside this trash bin!"
2 - "Once you have the data you bin it!"
3 - "My evidence is thiiiiis big!"
4 - "Sim-sala-bin - and here's your likelihood minimum!"
Join the fun and suggest a caption in the comments thread!
Accelerating Dark Matter Searches With Machine Learning
Comments