Introducing MLflow for End-to-End Machine Learning on Databricks

Solving a data science problem is about more than making a model. It entails data cleaning, exploration, modeling and tuning, production deployment, and workflows governing each of these steps. In this simple example, we'll take a look at how health data can be used to predict life expectancy. It will start with data engineering in Apache Spark, data exploration, model tuning and logging with hyperopt and MLflow. It will continue with examples of how the model registry governs model promotion, and simple deployment to production with MLflow as a job or dashboard.
About:
Databricks provides a unified data analytics platform, powered by Apache Spark™, that accelerates innovation by unifying data science, engineering and business.
Read more here: databricks.com/product/unifie...
Connect with us:
Website: databricks.com
Facebook: / databricksinc
Twitter: / databricks
LinkedIn: / databricks
Instagram: / databricksinc Databricks is proud to announce that Gartner has named us a Leader in both the 2021 Magic Quadrant for Cloud Database Management Systems and the 2021 Magic Quadrant for Data Science and Machine Learning Platforms. Download the reports here. databricks.com/databricks-nam...

[@chrisstephenson9890]

Excellent coverage of the capabilities available in Spark/Databricks generally and MLflow in particular.

[@rakesh7464]

Is there a repository on which we can find this code? Excellent content!

[@erv993]

Excellent workshop! thanks for sharing!

[@sandipmohanty8598]

very well explained and excellent content. got to know about hyperopt & shap packages and their usefulness. Super like

[@BernhardSchlegel]

Great content! I may have missed it: Let's say I have a model that requires my features to be transformed before handing them over to the predictor. Is it possible to include that in the mlfow model container? Or is this something that is required to happen, e.g., on the client if I'm serving using a REST API? Thanks.

[@huqianghui7625]

Where can I get the notebook file？

[@popalex]

Please add a repository location for the examples in this video.

[@quasenerd5476]

Great content! Is the notebook available?

[@joaoramos254]

very good stuff

[@ben-salem]

Could you please share the notebooks? great content

[@TitoDiaz6996]

Good video!

[@charlesaydin2966]

Thanks for the content! Question: At about 22 minute 4 sec mark, you say "when we log this..." and then click on something to get a different view. What do you click on there?

[@sumitbhalla2321]

Is there any api code snippets to enable model serving? we want to automate enable model serving. Please help. thank

[@meghanavasavada96]

Excellent Content !! Could you please share the notebooks and data source? Thanks !!

[@sandipmohanty8598]

please share the git hub code or notebook if any

[@FirstNameLastName-fv4eu]

It is very unfortunate to see the whole ML/AI world learned nothing from Statistical modeling which has been happening for the last 50-60 yrs. Maybe this is what happens when money becomes too cheap, and investment becomes too easy to achieve. This video is a perfect example of OVERENGINEERING!! The dev cost is so high it is practically impossible to earn any money by selling it and becoming cash positive. The bubble is waiting to kill the industry BADly.