KIPAC Stats & ML Journal Club

Everything you wanted to know (and many things you didn't) about Gaussian Processes

-
Location

Campus, PAB 232

Speaker
Sean McLaughlin

Join us Friday Feb 14th at 3pm in PAB 232 on campus (and on zoom ) for the next meeting of the Stats and ML Journal Club. This week, I (Sean McLaughlin) will be leading an introduction on Gaussian Processes. See you all then!

Title: Everything you wanted to know (and many things you didn't) about Gaussian Processes

An interpretable machine learning framework for dark matter halo formation

-
Location

Campus, PAB 232

Speaker
Ethan Nadler

Join us Friday Jan 17th at 3pm in PAB 232 on campus (and on zoom) for the first meeting of the Stats and ML Journal Club this quarter. This week, Ethan Nadler will be leading a discussion on models for interpretable halo formation. See you then!

 

World Models

-
Location

Campus, PAB 232 *note change in location*

Speaker
Sean McLaughlin

Join us Friday Jan 10th at 3pm in PAB 232 on campus (and on zoom) for the first meeting of the Stats and ML Journal Club this quarter. This week, I will be leading a discussion on World Models, an interesting approach to training RL agents using an emulator-like approach. Find a link to the paper (with great interactive visuals) below.

Approximating photo-z PDFs for large surveys

-
Location

SLAC, Kavli 2nd Floor Conf Room *note change in location on this date*

Speaker
Claire Hebert

Join us Friday Nov 15th at 3pm in the SLAC 2nd floor Fishbowl (and on zoom) for the Stats & ML Journal Club. 

This week, Claire Hebert will lead a discussion on optimally storing complex photo-z PDFs for galaxy surveys. Find the paper and abstract below.

 

Relational inductive biases, deep learning, and graph network

-
Location

SLAC, 3rd Floor Conference Room

Speaker
Lawrence Peirson (Stanford)

Join us Friday Nov 1st at 3pm in the SLAC 3rd Floor Conference Room (and on zoom) for the Stats & ML Journal Club. 

This week, Lawrence Peirson will lead a discussion on Graph Networks, more general extensions of conventional Neural Networks; should be very interesting! See the relevant paper and abstract below.

Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning

-
Location

SLAC, Kavli 3rd Floor Conf. Room

Speaker
Sebastian Wagner-Carena

Join us Friday October 25th at 3pm in the SLAC 3rd Floor Conference Room (and on zoom) for the Stats & ML Journal Club.

This week, Sebastian Wagner-Carena will lead a discussion on likelihood-free inference in strong lens modeling using CNNs. Find the paper and abstract below