Udemy

Solve Kaggle's OpenVaccine Challenge w/ Kubeflow and MLOps

Data Science, Kubeflow, Kale and MLOps come together in this course based on the Kaggle OpenVaccine Challenge;
Free tutorial
Rating: 5.0 out of 5 (3 ratings)
380 students
30min of on-demand video
English
English [Auto]

Articulate the relationship between the Kaggle OpenVaccine Competition and Kubeflow.
Outline the stages of MLOps and explain the value of Kubeflow as it pertains to MLOps.
Use Jupyter Notebooks in Kubeflow to review the Kaggle OpenVaccine Problem Solution.
Define Kubeflow Pipeline using Kale and Jupyter Notebooks hosted on Kubeflow Clusters.
Use Katib to perform Hyperparameter Tuning with Kubeflow Pipelines.
Load Kubeflow Pipeline Snapshots in new Notebook Servers to restore previous state.
Serve the ideal model from a Jupyter Notebook.
Articulate how Machine Learning technologies come together to support MLOps.

Requirements

  • Familiarity with popular Data Science concepts

Description

The Kaggle OpenVaccine problem is a popular Data Science topic. In this course, you will explore how to solve this problem with Kubeflow and Kale. In addition, you’ll learn how the work you are doing is the foundation for an effective and self-sustainable MLOps culture and platform solution that you can undertake at your enterprise.


This course is presented as a series of hands-on articles where you will learn about Kaggle, Data Science, and MLOps while using the Kubeflow platform with Kale to compile and run Kubeflow Pipelines. The overall time commitment is about 1 to 1.5 hours.


Specifically in this course, you will:

  • Learn about Kaggle.

  • Learn about Kubeflow.

  • Learn about MLOps.

  • Use Jupyter Notebooks in Kubeflow to review the Kaggle OpenVaccine Problem Solution.

  • Use Kale to convert a Jupyter Notebook into a Kubeflow Pipeline.

  • Use Katib to perform Hyperparameter Tuning on the ideal OpenVaccine model.

  • Load the Kubeflow Pipeline Snapshots in new Notebook Servers.

  • Serve the ideal OpenVaccine model from a Jupyter Notebook.

  • Relate the activities in this course back to the core tenets of MLOps.

Requirements: We assume that you have familiarity with popular Data Science concepts and have used some of these philosophies in practice.

Instructor-Led Option: If you would prefer to take the course live, this course is available on a monthly basis with an instructor. If this is your preference, navigate and sign up on the Arrikto events page.    

Who this course is for:

  • Anyone interested in learning more about Kaggle, Kubeflow and / or MLOps

Instructors

Educating the masses on Cloud Technologies
Alexander Aidun
  • 5.0 Instructor Rating
  • 3 Reviews
  • 380 Students
  • 1 Course

I am a Cornell Engineering Graduate with a degree in Information Systems, Science, and Technology. I focus my career on empowering people to be successful with emerging technologies. Through a combination of concise storytelling and hands-on activity, I am able to teach people new skills to advance their own careers.


Currently, I am working at Arrikto as the Director, of Education and we are pioneering MLOps on Kubeflow. This cutting-edge approach to the Model Development Life Cycle and Machine Learning Ops is the foundation for the future of Enterprise AI. I am eager to share this new technology with the Udemy community! 

Media lead at Arrikto
Ben Reutter
  • 5.0 Instructor Rating
  • 3 Reviews
  • 380 Students
  • 1 Course

I'm a multimedia developer that loves creating content for tech companies. At Arrikto I'm trying to share our story of how we are the leading force for MLOPs.

Arrikto’s Enterprise Kubeflow distribution is a complete MLOps platform that reduces costs, while accelerating the delivery of scalable models from laptop to production.

Top companies trust Udemy

Get your team access to Udemy's top 19,000+ courses