5 can’t-miss sessions about machine learning at Google Cloud NEXT ‘17

If you’re interested in the future of cloud computing or are a current/aspiring Google Cloud user (or both), attending Google Cloud NEXT ‘17 (March 8-10 in San Francisco) could become a new annual ritual for you. With hundreds of codelabs, bootcamps and breakout sessions, we’ve designed that conference to become a dream destination for the cloud computing ecosystem.

Although Google Cloud Next ‘17 has several themes befitting this multidimensional subject, machine learning is a first among equals. Machine learning has long been a keystone in Google’s own infrastructure, and recently, mainstream interest in it has grown rapidly. That process has accelerated with the emergence of the open source TensorFlow framework as a standard for deep-learning model development (it’s currently one of the most active projects on GitHub), and the availability of Google Cloud Machine Learning APIs (for those who prefer pre-built, pre-trained models). That strong interest we see among the growing machine-learning community generally, and Google Cloud customers specifically, is reflected in the conference agenda.

There are many options from which to choose, but here are a few examples of sessions (listed in chronological order) that I recommend:

• BigQuery and Cloud Machine Learning: advancing large-scale neural network predictions (March 8, 1:20pm - Kaz Sato, Developer Advocate, Google Cloud)
  Bringing access to machine learning into the Google BigQuery workflow will “democratize” machine learning like never before. This session demonstrates how to combine Cloud Machine Learning and BigQuery to realize this vision.
• TensorFlow and deep learning without a PhD, Part 1 (March 8, 4pm - Martin Görner, Developer Advocate, Google Cloud)
  For those interested in building customized models with TensorFlow, lack of deep-math background needn’t be a blocker. In this session (first of two parts), attendees will learn how to build a simple model as well as some of the tricks of the trade in neural-net design.
• Machine learning as an API: accessing pre-trained ML models with one API call (March 9, 11:20am - Sara Robinson, Developer Advocate, Google Cloud)
  Here Sara offers a code-level overview of how to bring pre-trained models into your apps via Google Cloud Machine Learning APIs (Google Cloud Vision API, Cloud Speech API, Cloud Natural Language API and Cloud Translation API); demo included. (For a deep dive into development with Cloud ML APIs, you should also consider participating in the “Serverless Machine Learning with Google Cloud ML” bootcamp" on March 6 or 7.)
• Reimagining human computer interaction with Cloud Speech API (March 9, 4pm - Dan Aharon, Product Manager, Google Cloud)
  Human-machine interaction is transforming user experiences. Here you’ll learn how the Google Cloud Speech API helps developers create those experiences without the need for model building or training.
• Preventing Overfishing with Machine Learning and Big Data Analytics​ (March 9, 5:20pm - David Kroodsma,
  Global Fishing Watch Research Manager, SkyTruth and Amy Unruh, Developer Programs Engineer, Google)
  Machine learning is often a stage in a broader data analytics continuum. In this session, attendees will learn how to use Google Cloud Dataflow to transform streaming data for distributed training of a neural net via Google Cloud Machine Learning (aka, TensorFlow-as-a-service).