Agenda (click below for the workshop agenda):
Presentations:
Introduction to Machine & Deep Learning
Frontiers for Deep Learning and Cancer - Eric Stahlberg, Frederick National Laboratory for Cancer Research:
Introduction to Deep Learning - Rick Stevens, Argonne National Laboratory:
Deep Learning: Perspectives from the NIH
Accelerating Image Analysis Workflows Using Deep Learning - Yanling Liu, Frederick National Laboratory for Cancer Research:
The Impact of Deep Learning on Radiology - Ronald Summers, Diagnostic Radiology Department, NIH Clinical Center:
Applying Deep Learning to Big Data: Perspectives from the Department of Energy with Applications in Health
Cellular Level Deep Learning - Fangfang Xia, Argonne National Laboratory:
Molecular Level Deep Learning - Brian Van Essen, Lawrence Livermore National Laboratory:
Population Level Deep Learning - Arvind Ramanathan, Oak Ridge National Laboratory:
Getting Started With CANDLE - Rick Stevens, Argonne National Laboratory:
Overview of CANDLE Layers: Workflows, Scripting, and Parallelization Strategies
Workflows - Justin Wozniak, Argonne National Laboratory:
Scripting - Tom Brettin, Argonne National Laboratory:
Parallelization Strategies - Brian Van Essen, Lawrence Livermore National Laboratory:
Preparing Data for Deep Learning
Arvind Ramanathan, Oak Ridge National Laboratory:
Fangfang Xia, Argonne National Laboratory:
Brian Van Essen, Lawrence Livermore National Laboratory:
Interactive Session with CANDLE Developers
Hyperparameter Optimization and Uncertainty Quantification - Justin Wozniak, Argonne National Laboratory:
Installing CANDLE into Other Environments - George Zaki, Frederick National Laboratory for Cancer Research:
Workshop Recordings:
Day 1 - Wednesday, February 21, 2018: CANDLE Workshop - Day 1 - Wednesday, February 21, 2018-20180221 1348-1.arf
Day 2 - Thursday, February 22, 2018: CANDLE Workshop - Day 2 - Thursday, February 22, 2018-20180222 1400-1.arf
**Note: The WebEx ARF player is required to playback the recording. Download the ARF player HERE: https://cbiit.webex.com/client/31.11.11/nbr2player.msi
2018 Workshop Presenters:
• Eric Stahlberg, Frederick National Laboratory for Cancer Research (eric.stahlberg@nih.gov)
• George Zaki, Frederick National Laboratory for Cancer Research (george.zaki@nih.gov)
• Yanling Liu, Frederick National Laboratory for Cancer Research (liuy5@mail.nih.gov)
• Ronald Summers, NIH Clinical Center, Diagnostic Radiology Department (RSummers@cc.nih.gov)
• Rick Stevens, Argonne National Laboratory (Stevens@anl.gov)
• Fangfang Xia, Argonne National Laboratory (fangfang@anl.gov)
• Brian Van Essen, Lawrence Livermore National Laboratory (vanessen1@llnl.gov)
• Arvind Ramanathan, Oak Ridge National Laboratory (ramanathana@ornl.gov)
• Justin Wozniak, Argonne National Laboratory (wozniak@mcs.anl.gov)
• Tom Brettin, Argonne National Laboratory (brettin@cels.anl.gov)
2017 CANDLE Workshop Artifacts:
LINK: 2017 CANDLE Workshop Documents and Presentations
Event Recap:
With over 150 attendees representing 22 NIH institutes, government collaborators, industry, and academia, the February CANDLE Workshop at NIH proved to be a truly engaging and successful event!
Throughout the two-day workshop, many different perspectives on machine and deep learning were explored along with their applications to cancer research and advancing precision oncology.
The first day included a general overview of machine and deep learning along with applications of deep learning in health from the NIH and Department of Energy perspectives.
The second day included hands-on training of deep learning along with tutorials of how to prepare data for deep learning and how to install the CANDLE deep learning framework into other environments.
Some of the common exploratory areas of interest included imaging analysis, next-generation sequencing, genomics and genetics, text analysis, and big data.
As most workshop participants did not have much previous experience with machine or deep learning, this event provided the opportunity for participants to progress from little or no understanding of ML/DL technologies, to having the ability to train deep learning models with their own data sets, and on their own computers and laptops!
As an outcome of this workshop, future events are already in the works which will aim to address the diverse set of challenges raised within the community, and to meet the increasing demands deep learning technology in present-day research.
