The 5th i-CoMSE Workshop: Machine Learning for Molecular Science

Description: This workshop will provide an overview of machine learning tools applied to study molecular systems with a focus on computational molecular science. It will feature a mix of lectures and hands-on exercises running machine learning algorithms with molecular simulations on national supercomputing resources. The workshop will include a session on diversity, equity, and inclusion aspects of computational sciences. Sessions will be taught in Software Carpentry style, with approximately equal time divided between lectures and hands-on programming exercises.

Registration: The application for the workshop can be accessed here

Preliminary Schedule:

  • Monday July 10th, 2023
    • Feature representation
    • Dimensionality reduction
    • Unsupervised learning
  • Tuesday, July 11th, 2023
    • Clustering and visualization
    • Regression and supervised learning
    • Training, testing, and validation
    • Simple regression models
    • D&I workshop
  • Wednesday, July 12th, 2023
    • Structured vs unstructured data
    • Regularized vs tree based methods
    • Interpretability
    • Workshop Outing
  • Thursday, July 13th, 2023
    • Deep learning
    • Advanced examples of ML applications to molecular science
    • ML on cyberinfrastructure
  • Friday, July 14th, 2023
    • Generative models and inverse design
    • General Q&A
    • Wrap up

Expected number of participants: 40

Equipment Needed: Students will need a laptop and have Jupyter notebooks installed. Computing resources for performing exercises will be provided in Bridges-2 at PSC.

Dormitory Check-In: July 9th, 2023 (starts at 1 pm) and Check out: July 14th 2023 (by noon)

Eligibility: Participation is limited to graduate students, undergraduate students (with junior or senior standing), postdoctoral, early career faculty, with some experience in molecular simulations and Python scripting. Applicants must be associated with an US based educational institution, national lab, or industry.

Registration Fee: None for academic participants

Financial Support: Financial support is available to cover housing (in university dormitory), meals, and parking cost for all non-local participants. A limited number of travel stipends (up to $500) are available to help broaden participation.

Organizers/Instructors:

  • Chris Bartel, University of Minnesota Twin Cities
  • Rose Cersonsky, University of Wisconsin-Madison
  • Eric Jankowski, Boise State University
  • Neeraj Rai, Mississippi State University
  • Sapna Sarupria, University of Minnesota Twin Cities
  • Jindal Shah, Oklahoma State University
  • Michael Shirts, University of Colorado, Boulder