GeoSmart Hackweek 2023

23 October - 27 October Seattle, WA
Application of Machine Learning in Hydrology and Cryosphere Science.

2023 Event Starts In:

Schedule

All times listed below are UTC -7 (Pacific Daylight Time). You might want to consult this Time Zone Map to figure out times in your location. Tutorials will be located in Alder Auditorium.

8:30 - 9:30

Intro and welcome

Getting to know each other setting the stage for our work together.

9:30 - 9:45

BREAK

9:45 - 10:45

CryoCloud and Jupyter Notebooks

Getting connected to our shared computational resources.

Tutorial Lead(s)

10:45 - 11:00

BREAK

11:00 - 12:00

Introduction to machine learning

An overview of machine learning and the primary stages of typical Machine Learning workflows including data preparation, model selection, training, and visualization.

Tutorial Lead(s)

12:00 - 13:00

LUNCH

13:00 - 15:00

Intro to Projects

Overview of projects for this hackweek and how we will gather in small groups. Choosing our project team for the week.

15:00 - 16:00

Project Work

8:30 - 9:00

Approaches to Working in Teams

Discussion of best practices for collaborative work. Reflecting on the opportunities and challenges of small group work.

Tutorial Lead(s)

9:00 - 10:00

ML methods for regression problems Part 1

Introducing neural networks

Tutorial Lead(s)

10:00 - 10:30

BREAK

10:30 - 12:00

Project Work

12:00 - 13:00

LUNCH

13:00 - 14:00

ML methods for regression problems Part 2

Introducing tree-based methods

Tutorial Lead(s)

14:00 - 16:00

Project Work

8:30 - 9:45

ML Workflow Management - GeoWEAVER

Guidance on managing reproducible ML workflows.

Tutorial Lead(s)

9:45 - 10:15

BREAK

10:15 - 12:00

Project work

12:00 - 13:00

LUNCH

13:00 - 13:30

Group photo

13:30 - 16:00

Project Work

8:30 - 9:30

ML for classification

Using ML methods for classification purposes.

Tutorial Lead(s)
Kehan Yang

9:30 - 10:00

BREAK

10:00 - 12:00

Project Work

12:00 - 13:00

LUNCH

13:00 - 15:00

Project Work

15:00 - 16:00

Machine Learning Using Google Earth Engine

Tutorial Lead(s)
Steve Greenberg

8:30 - 9:00

Ideas for Community Building and Open Science

Setting a course for continued engagement with CUAHSI and the ML geoscience community of practice.

Tutorial Lead(s)
Tony Castronova

9:00 - 10:00

Project wrap-up

10:00 - 10:30

BREAK / FOCUS GROUP

10:30 - 12:00

Sharing Project Outcomes Part 1

An opportunity to show what we accomplished during the week.

12:00 - 13:00

LUNCH

13:00 - 13:30

Hackweek Survey

Complete a survey providing us with feedback for improving our program.

Tutorial Lead(s)

13:30 - 15:30

Sharing Project Outcomes Part 2

An opportunity to show what we accomplished during the week.

15:30 - 16:00

Closing

Tutorial Lead(s)

Meet the team

The people on this page have helped organize the hackweek. You'll find a few specializations listed per person if you're wondering who to reach out to during the event!
Nicoleta Cristea
Research Scientist
Lead Hackweek Organizer badge
Read more →
Anthony Arendt
Senior Data Science Fellow
Lead Hackweek Organizer badge
Read more →
Mark Welden-Smith
Program Manager
Program Manager badge
Read more →
Charley Haley
Participant Interaction and Collaboration Architect
Lead Community Building badge
Read more →
Scott Henderson
Research Scientist
Lead of Technology badge Tutorial Lead badge
Read more →
Don Setiawan
Research Software Engineer
Technology Specialist badge
Read more →
Naomi Alterman
Education Consultant
Education Consultant badge
Read more →
Irene Garousi-Nejad
Research Scientist
Hackweek Helper badge Project Consultant badge
Read more →
Aji John
Postdoc
Tutorial Lead badge
Read more →
Friedrich Knuth
PhD Student
Hackweek Organizer badge
Read more →
Steven Pestana
Research Scientist
Hackweek Organizer badge
Read more →
Morgan Sanger
PhD Student
Hackweek Helper badge
Read more →
Ziheng Sun
Research Scientist
Tutorial Lead badge
Read more →
Tasha Snow
Postdoctoral Researcher
CryoCloud Lead badge Tutorial Lead badge
Read more →
Ann Nykamp
Grants Manager
Grants Manager badge
Read more →
Ibrahim Olalekan Alabi
Graduate Student
Tutorial Lead badge
Read more →
Ryan Johnson
AI/ML Research Scientist
Tutorial Lead badge
Read more →
Valentina Staneva
Data Scientist
Tutorial Lead badge Hackweek Helper badge
Read more →

About GeoSmart Hackweek

Hackweeks are participant-driven events that strive to create welcoming spaces for participants to learn new things, build community and gain hands-on experience with collaboration and team science.

The 2023 GeoSmart Hackweek will focus primarily on project work, with an emphasis on applications in Hydrology and Cryosphere science. Project ideas will be shared in advance. There will be space for new ideas to emerge during the Hackweek based on participant engagement. We will provide about 5 hours of data science tutorials spread over the week, including space for participant-led tutorials.

During the week, participants will have the opportunity to collaborate with their peers, share ideas, and work on projects leading to exciting results and discoveries. This event is open to all experience levels in machine learning knowledge, so whether you're a seasoned pro or just starting out, you're welcome to join. However, to benefit most from the event, prior knowledge of Python programming and data handling using common Python packages (pandas, xarray, etc.) is desired. See the event Jupyter book for more details.

Preliminary project ideas include streamflow prediction from SAR-derived snowmelt timing or snow data, predicting snow water equivalent with machine learning, glacier dh/dt from DEMs using geospatial time series analysis, derivation of snow covered areas from satellite imagery, derivation of snow depth from SAR backscatter and lidar-derived snow data, predicting river discharge from seismic waves and others! Join one of these projects or pitch your own project idea at the event!

    Brainstorming sessions : participants can join an existing project or come up with ideas for projects that can be implemented using machine learning.
    Tutorials : learn about common machine learning workflows, computational environments, reproducibility, and workflow management.
    Data preparation : explore datasets to identify and engineer relevant variables that can be used to build machine learning models.
    Models : work on building machine learning models using popular libraries such as TensorFlow, PyTorch, or scikit-learn.
    Model validation and optimization : validate models using cross-validation and other techniques to ensure that models are robust and accurate; fine-tuning hyperparameters, using feature engineering techniques, or other methods.
    Presentations : participants can share the results from projects to receive feedback from their peers.
    Networking : facilitated opportunities for networking and community building will be provided.

Our Sponsors

This event was made possible by the National Science Foundation (Awards #1829585, #2117834) and the eScience Institute in collaboration with CUAHSI and ESIP. Cloud computing infrastructure provided by CryoCloud.
eScience Institute
National Science Foundation
CUAHSI
ESIP
CryoCloud