MLclub.net

 
Machine Learning Club
An online discussion on Machine Learning
for astrophysicists


Organizers:

François Lanusse 
(CEA Saclay)

Brice Ménard 
(Johns Hopkins)

Marc Huertas Company 
(IAC, Paris Observatory)

J. Xavier Prochaska 
(UC Santa Cruz)

Meeting Logistics:

Academic year 2021-2022

2021-05-26

Debate: ML & stellar spectroscopy

Panel:

Yuan-Sen Ting
(IAS)

Kim Venn
(Victoria)

David W. Hogg
(NYU) 

Keith Hawkins
(U. Texas)

Moderator: J. X. Prochaska [slides, recording]

2021-05-12

Debate: ML Uncertainty Quantification for Scientific Applications

Panel:

Laurence Perreault-Levasseur
(Universite de Montreal)

Tilman Plehn
(Heidelberg University)

Tom Charnock
(Freelance consultant)

Andrew Gordon Wilson

(NYU CIMS & CDS)

Moderator: F. Lanusse [Recording]

2021-04-28

Debate: The ML impact in cosmology


Panel:

Bhuvnesh Jain
(UPenn)

Uros Seljak
(Berkeley)

Roberto Trotta
(Imperial College) 

Hiranya Peiris
(UCL)

Moderator: B. Ménard [Recording]

2021-04-14

Debate: Outliers: How do we discover them with ML and what are they good for?

Panel:

Ben Nachman
(LBNL)

Dalya Baron
(Tel Aviv University)

Hannah Kenner
(University of Maryland) 

Chris Lintott
(Oxford University)

Moderator: X. Prochaska [Recording]

2021-03-31

Debate: How to study galaxy morphology?

Panel:

Simon White
(Max Planck Institute)

Helena Dominguez-Sanchez
(Barcelona)

Josh Peek
(Space Telescope Science Institute)

Sandy Faber
(UCSC)

Moderator: M. Huertas-Company [Recording]

2021-03-17

Debate: Will ML solve Photometric Redshifts?

Panel:

Gary Bernstein
(UPenn)

Olivier Ilbert 
(LAM, Marseille)

Alex Malz
(GCCL@RUB) 

Emmanuel Bertin
(IAP, Paris)

Moderator: F. Lanusse. [Slides] [Recording]

2021-03-03

Debate:  How should ML penetrate the natural sciences? Do we need ML institutes?

Panel:

David Spergel
(Flatiron Institute)

Julia Kempe
(NYU Center for data science)

Alex Szalay
(Johns Hopkins)

J. Xavier Prochaska
(UC Santa Cruz) 

Moderator: B. Ménard. [Slides] [Recording]

2021-02-03

Debate: What will ML do (or not) for the Rubin Observatory project (LSST)?

Panel:

Robert Lupton
(Princeton)

Josh Bloom
(Berkeley)

Vanessa Boehm
(Berkeley) 

Peter Melchior
(Princeton)

Moderator: B. Ménard, Recording

2021-01-20

Self-Supervised Representation Learning for Astronomical Images Slides

Presentation by George Stein, UC Berkeley

Academic year 2020-2021

2020-12-16

Debate: What is Deep Learning Teaching Astronomy?

Moderator: Alexie Leauthaud     Panel: The ML Club organizers

https://ucsc.zoom.us/j/98445420245 

2020-12-02

Siamese Neural Networks Learn Symmetry Invariants and Conserved Quantities

Speaker: Sebastian Wetzel (Perimeter Institute for Theoretical Physics)

2020-11-18

Denoising Score Matching for Uncertainty Quantification in Inverse Problems: Application to gravitational lensing and Magnetic Resonance Imaging

Speakers: Benjamin Remy, Zaccharie Ramzi (CEA Paris-Saclay)

arXiv paper, arXiv paper, recording

2020-10-21

Learning effective physical laws for generating cosmological hydrodynamics with Lagrangian Deep Learning | arXiv paper

Speaker: Biwei Dai (Berkeley)

2020-10-07

A new vocabulary for patterns | arXiv paper

Speaker: Sihao Cheng (Johns Hopkins Univ.)

Beyond the Hubble Sequence - Exploring galaxy morphology with unsupervised machine learning | arXiv paper

Speaker: Ting-Yun Cheng (Nottingham)

2020-09-23

Neural Scaling Laws and GPT-3

Speaker: Jared Kaplan (Johns Hopkins / OpenAI)

Paper: https://arxiv.org/abs/2005.14165

Lecture notes on Machine Learning for Physicists

2020-09-09

Feature Extraction on Synthetic Black Hole Images

Speaker: Joshua Yao-Yu Lin (U. Illinois)

Paper:  https://arxiv.org/abs/2007.00794

Anomaly Detection in Hyper Suprime-Cam Images with Generative Adversarial Networks

Speaker: Kate Storey-Fisher (NYU)
Recording

2020-08-26

Discovering Symbolic Models from Deep Learning with Inductive Biases
Speaker: Miles Cranmer
Paper:
https://arxiv.org/abs/2006.11287  |   Recording

Academic year 2019-2020

2020-07-29

Holiday break

2020-07-15

Flows for simultaneous manifold learning and density estimation #notagan

Speaker: Johann Brehmer (NYU; https://johannbrehmer.github.io/)

This work is based primarily on this paper: https://arxiv.org/abs/2003.13913

2020-07-01

Learning maths from example with deep language models

Speakers: François Charton (FAIR), Amaury Hayat (Paristech, Rutgers)  
The work is based on these 2 papers:

https://arxiv.org/pdf/1912.01412.pdf , https://arxiv.org/pdf/2006.06462.pdf

2020-06-03

A Deep Dive into CapsNets

Speaker: J. Xavier Prochaska (AAII)

2020-05-20

SimCLR: A Simple Framework for Contrastive Learning of Visual Representations

Speaker: Ting Chen (Google Research)

2020-05-06

Quasar continua predictions with neural spline flows

Speaker:  David Reiman (UCSC)
QuasarNET
Speaker:  / James Farr (UCL)

2020-04-08

Neural Networks with Euclidean Symmetry for Physical Sciences
Jointly held with the
AAII Seminar;  at 12pm PT
Speaker: Tess Schmidt (LBNL)
Zoom link

2020-03-11

Statistical inference of dark matter substructure from strong gravitational lenses without a likelihood

Speaker: Siddarth Mishra-Sharma (NYU)

2020-02-26

Astronomical images as a playground to understand OoD behavior of generative models 
speaker: Lorenzo Zanisi (Southampton)
Regularizing Normalizing Flows for Out-of-Distribution Detection

speaker: Vanessa Boehm (Berkeley)

2020-02-12

Likelihood Ratios for Out-of-Distribution Detection
Speaker: Balaji Lakshminarayanan (DeepMind)

Paper links are found here:  http://www.gatsby.ucl.ac.uk/~balaji/

2020-01-29

Full-Gradient Representation for Neural Network Visualization

Speaker: Suraj Srinivas (EPFL)

2020-01-15

Lesson from bringing software engineering to machine learning

Speakers: Pippin Lee, Cole Clifford (Dessa)

2019-12-19

Remote sensing & ML | slides

Speaker: Hannah Kerner (UMD)

2019-12-05

An Overview of Graph Networks and Physical Simulations

Speaker: Jonathan Godwin (DeepMind)

2019-11-21

More on deep probabilistic learning

Speaker: Francois Lanusse (Berkeley)

Bayesian models and active learning

Speaker: Mike Walmsley (Oxford)

Meeting docs

2019-11-07

Deep probabilistic learning

Speaker: Francois Lanusse (Berkeley)

2019-10-24

Transformers

Speaker: David Reiman (UCSC)

Academic year 2018-2019

2019-06-19

RE:MARS conference debriefing

Speaker: Xavier Prochaska

2019-05-22

--

2019-05-08

Applications of deep generative models from emulating complex physical processes to solving inverse problems

Speaker: Francois Lanusse (Berkeley)

2019-04-24

Pixel Level Morphological Classification using Semantic Segmentation

Speaker: Ryan H.

Galaxy deblending

Speaker: Alexandre Boucaud

2019-05-10

Activation Atlases

Speaker: Brice Ménard (Johns Hopkins)

2019-03-13

Graph neural networks: https://arxiv.org/abs/1812.08434

Speaker: Marc Huertas Company (Paris Observatory)

2019-02-27

Deep learning detection of transients

Speaker: Dalya Baron (Tel Aviv Univ.)

2019-02-13

Style transfer

Speaker: Xavier Prochaska  (UCSC)

2019-12-19

The scattering transform

Speaker: Brice Ménard (Johns Hopkins)

2018-12-05

Integrated gradients on DLAs

Speaker: Xavier Prochaska (UCSC)

2018-10-24

DC update -- Speaker: Cheng

Latest fun with GANs -- Speaker: Reiman

Image classification of DES -- Speaker: Marc

2018-10-10

Updates on his U-Net -- Zheng

Latest fun with GAN -- Reiman

Quantifying classification errors with random forests -- Dalya Baron

Identifying informative pixels with neural nets. From saliency to Integrated gradients -- Josh Peek

2018-09-26

Summary of Rework Deep Learning Summit

Galaxy deblending

2018-09-12

Second discussion

2018-08-29

First discussion