Earth System Data Science in the Cloud

Session Overview

  • Welcome
  • Interim Check-In
  • Data Driven Science
  • What is Data Science?
  • Where We Are/Context
  • Course Goals and Objectives
  • Module Goals and Objectives
  • Course Logistics

Welcome To Module 4!

Interim Check-In


How did you apply Module 3 to your work?


What did you explore?

Where are you in your team environment?

This Module

Data Product Development

Data Driven Science

Schleder et al. 2019 DOI: 10.1088/2515-7639/ab084b

Data Science

Where We Are/Context

Course Goals & Objectives

  1. Make the Impossible Possible
  2. 10x Performance

Course Goals & Objectives

  • Conversant and practiced in developing cloud based earth system data science workflows.

  • Comfortable developing data science products.

  • Comfortable and practiced in working effectively on interdisciplinary teams.

  • Able to rapidly pick up new skills, tools, and techniques.

Principles & Practices

Module Goals & Objectives

By the end of this module, you will be familiar with and conversant in the following areas:

  • Visualizing Earth Systems data at scale.
  • Production Machine Learning in the cloud.
  • End-to-End Pipeline Development.
  • Effective and Efficient Publication Development.

Module Goals & Objectives

Specifically by the end of the course, you will have accomplished the following:

  • Polished a pipeline for dataset development
  • Developed a pipeline with multiple ML models.
  • Developed visualizations of larger than memory data
  • Effectively documented your code
  • Launched a machine learning model using a REST API
  • Developed submission quality reports using LaTeX and Overleaf
  • Communicated the results of your team projects in formal presentations.

Module Outline

Days:

  1. Managing Multiple Models
  2. Ethics & Intro to Deep Learning
  3. Data Viz & Deep Learning (& Guest Lecture!)
  4. Production ML
  5. Team Presentations and Next Steps

Team Project Outline

Days:

1-3. ML & Project Development

  1. Presentation Practice
  2. Presentations

Team Project Deliverables

Reference Templates. More details during team time.

Presentation

  • 10 minutes (no more, can be less)
  • Team Name
  • Results
  • Next Steps

Report

  • LaTeX/Overleaf
  • Document your Modeling Approach to Date
  • This will turn into methods and Results section of paper

Session Overview

  • Welcome
  • What is Data Science?
  • What is Earth System Data Science in the Cloud?
  • Course Goals and Objectives
  • Module Goals and Objectives
  • Course Logistics

Course Logistics

Strategies for Success

This is a lot of information.

  • You would not be here if you could not handle it.
  • Be present.
  • You will not understand everything the first time. That is OK!
  • Keep a Journal of topics to return to and explore more
  • You will see each topic/idea at least 3 times on separate days
  • Ask questions
  • Invest the time now…

Final Note

We made this course for you! We want your feedback!

Please reach out anytime on Slack or at dwillett@cicsnc.org & ggraham@cicsnc.org.

Team Pre-Project Assessment

Production Assessment

Earth System Data Science in the Cloud