Organizers:
Rama Vasudevan, Stephen Jesse
CNMS User Meeting Workshop 2018: Machine learning in Materials Science with Python
Welcome! This workshop will go over machine learning methods for materials science with python. The full agenda is published below. For remote participants, please click https://bluejeans.com/1766282540
The workshop will be conducted on August 15, 2018 at ORNL. The livestream begins at 9AM EDT.
To get the most out of this workshop, you may want to work along with us on your own laptop. If you do, we highly recommend that you:
1. Download and install the following:
• Anaconda5.2 for python 3.6 - https://www.anaconda.com/download/
• Pycroscopy- https://pycroscopy.github.io/pycroscopy/install.html
• HDFview- https://support.hdfgroup.org/products/java/release/download.html
2. Familiarize yourself with basics concepts.
Links to quick and helpful tutorials are available here - https://pycroscopy.github.io/pyUSID/external_guides.html.
Topics include:
• Basic python usage
• Jupyter notebooks for running code
• Numpy for numeric operations
Agenda
Time | Details |
---|---|
8:00-9:00 AM | Setup and installation |
9:00-9:30 AM | Introduction to machine learning in python |
9:30-10:15 AM | Linear unmixing methods for spectral analysis |
10:15-10:30AM | Coffee Break |
10:30-12:00 PM | Nonlinear unmixing: overview and applications |
12:00-1:00 PM | Lunch on your own |
1:00-2:00 PM | Traditional classification methods |
2:00-2:45 PM | Clustering methods applied to 4D STEM data |
2:45-3:00 PM | Break |
3:00-4:30 PM | Neural networks and deep learning for microscopy images |
4:30-5:00 PM | Wrap up/Discussion/Summary |
4D STEM dataset download link: https://ftp.ornl.gov/ftp_out/uP10oaHfq0g3/4D%20STEM%2020120212_21_GBb.h5