Difference between revisions of "AI tutorials"

From GlueXWiki
Jump to: navigation, search
(Created page with "= Purpose = * Provide an overview of the methods to search for exotic mesons * Help students and new collaborators get started with the technical steps necessary for a physic...")
 
(→‎Purpose)
Line 1: Line 1:
 
= Purpose =
 
= Purpose =
  
−
* Provide an overview of the methods to search for exotic mesons
+
The overall topic is: „Best practice for AI in nuclear physics applications“. We wish to cover the following items via plenary talks and/or interactive tutorials:
−
* Help students and new collaborators get started with the technical steps necessary for a physics analysis
+
 
−
* All tutorials should be explained on the basis on a fictitious analysis of the channel '''&gamma;p&rarr;&eta;&pi;<sup>0</sup>p'''
+
* Feature engineering --> Feature normalization, correlation coefficients, feature selection, etc.
−
&nbsp;
+
* Overfitting --> Dropout layers, weight regularization, etc.
 +
* Performance evaluation --> ROC-Curve, confusion matrix, accuracy, loss curves,...
 +
* Visualization --> How to properly present the performance of a model ? What are "good" diagnostic plots ?
 +
* Model deployment --> How to use a model within the GlueX analysis software
 +
* Data fed into models --> What data sets are used ? Numerical, vs. Images, raw data vs. clean data,...
 +
* Tools made available by the data science department
 +
* Optional, depending on time: HPO --> Tune the parameters of your model
 +
* Optional, depending on time: Bookkeeping of models via MLFlow
  
 
= Location and Time =
 
= Location and Time =

Revision as of 15:13, 14 January 2025

Purpose

The overall topic is: „Best practice for AI in nuclear physics applications“. We wish to cover the following items via plenary talks and/or interactive tutorials:

  • Feature engineering --> Feature normalization, correlation coefficients, feature selection, etc.
  • Overfitting --> Dropout layers, weight regularization, etc.
  • Performance evaluation --> ROC-Curve, confusion matrix, accuracy, loss curves,...
  • Visualization --> How to properly present the performance of a model ? What are "good" diagnostic plots ?
  • Model deployment --> How to use a model within the GlueX analysis software
  • Data fed into models --> What data sets are used ? Numerical, vs. Images, raw data vs. clean data,...
  • Tools made available by the data science department
  • Optional, depending on time: HPO --> Tune the parameters of your model
  • Optional, depending on time: Bookkeeping of models via MLFlow

Location and Time

The workshop will take place at:

DATES: May 23, 2022

LOCATION: CEBAF Center F113

 

Remote Participation

Join ZoomGov Meeting https://jlab-org.zoomgov.com/j/1601990060 (Click "Expand" to the right for more details -->):

Meeting ID: 160 199 0060 Passcode: 8394

One tap mobile 

+16692545252,,1608686187# US (San Jose)
+16468287666,,1608686187# US (New York)

Dial by your location 

+1 669 254 5252 US (San Jose)
+1 646 828 7666 US (New York)
+1 669 216 1590 US (San Jose)
+1 551 285 1373 US
   833 568 8864 US Toll-free

Meeting ID: 160 868 6187

Find your local number: https://jlab-org.zoomgov.com/u/akv0IBrQk

Join by SIP 

11601990060@sip.zoomgov.com

Join by H.323 (Polycom) 

161.199.138.10 (US West)
161.199.136.10 (US East)
Meeting ID: 160 199 0060
Passcode: 8394

 

References

It is not necessary to have gone through the previous workshop talks, but you will probably benefit if you do.

 

Workshop Software

The official software and all scripts used during the tutorial can be found on the JLab CUE: 

/group/halld/Software/gluex_workshops/tutorial_2022

Most software and scripts are available by cloning this GitHub repository and going down to the tutorial_2022 directory: 

https://github.com/JeffersonLab/gluex_workshops

Large data files related to the workshop can be found in: 

/work/halld/gluex_workshop_data/tutorial_2022

Agenda

Analysis Tutorial and Workshop

 

Registration

Please add your name to the list of attendees below. No formal registration or registration fee is required.

Name Home Institution Level Participate at JLab
Alex Austregesilo JLab Staff Yes
Matt Shepherd Indiana U. Prof. Yes
Edmundo Barriga FSU Grad. Student Yes
Jason Barlow FSU Grad Student Yes
Peter Pauli U. of Glasgow PostDoc No
Karthik Suresh U. of Regina PhD Student No
Varun Neelamana U. of Regina PhD Student No
Susan Schadmand GSI-FFN Staff Yes
Bhesha Devkota MSU Grad Student Yes
Lydia Lorenti W&M Grad Student Yes
Churamani Paudel FIU Grad Student Yes
Justin Stevens W&M Prof. Yes
Mariana Khachatryan FIU Postdoc Yes
Alison LaDuke CMU Grad Student Yes
Lawrence Ng FSU Grad Student Yes
Chandra Akondi FSU Postdoc No
Sean Dobbs FSU Prof. Yes
Donavan Ebersole FSU Grad Student Yes
Gabriel A. Rodriguez Linera FSU Grad Student Yes
Zachary Baldwin CMU Grad Student Yes
Jiawei Guo CMU Grad Student Yes
Beni Zihlmann JLAB Staff Yes
Tim Kolar TAU Postdoc No
Kevin Saldana Indiana U. Grad Student No
Jon Zarling Regina Postdoc Yes
Bo Yu Duke Grad Student No
Phoebe Sharp GWU Grad Student No
Rupesh Dotel FIU Grad Student No
Simon Taylor JLab Staff Yes
Torri Jeske JLab Postdoc Yes
Ryan Mitchell Indiana U. Scientist Yes
Logan Earnest GWU Undergraduate No
Albert Fabrizi UMass Grad Student No
Rebecca Barsotti Indiana U. Grad Student Yes
Kevin Scheuer W&M Grad Student Yes
Andrew Schick UMass Grad Student No
Elton Smith JLab Staff Yes
Nilanga Wickramaarachchi CUA Postdoc No
Daniel Lersch FSU Postdoc Yes
Joerg Reinhold FIU Prof. Yes
Viviana Arroyave FIU Grad Student Yes
Tolga Erbora FIU Grad Student Yes
Malte Albrecht Indiana U. Postdoc Yes
James McIntyre UConn Grad Student No
Keigo Mizutani JLab Postdoc Yes
Dene Hoffman CMU Grad Student Yes
Eric Habjan UConn Undergraduate No
Igor Strakovsky GW Prof. Yes

 

Recording

  • Links to video recordings hosted on YouTube are given in the agenda above.
  • A video recording of this workshop can be found on the JLab CUE in: /cache/halld/workshops/physics_workshop_2022/recordings
  • If the files are not in that directory, they can be retrieved from tape with the following command: jcache get /mss/halld/workshops/physics_workshop_2022/recordings/*
Retrieved from "https://halldweb.jlab.org/wiki/index.php?title=AI_tutorials&oldid=128113"