Within a couple of years of its release as an open-source machine learning and deep learning framework, TensorFlow has seen an amazing rate of adoption. Consider the number of stars on its github page: 98+K; look at the number of contributors: 1500+; and observe its growing penetration and pervasiveness in verticals: from medical imaging to gaming; from computer vision to voice recognition and natural language processing.
So it’s no surprise to me as a Program Chair of Spark + AI Summit to see high-caliber technical talks as part of the new tracks related to Deep Learning Techniques. In this blog, we highlight a few talks that caught our eye, in their promise and potential. It always helps to have some navigational guidance if you are new to the summit or technology.
Consider how Salesforce is using Apache Spark and TensorFlow to monitor customer activities in real-time to surface insights. Messrs Alexis Roos and Wenhao Liu will share in their talk, Deep Learning for Natural Language Processing Using Apache Spark and TensorFlow, how to build an LSTM classifier using the TensorFlow framework and combine the deep learning apparatus of TensorFlow with the distributed data processing power of Apache Spark. Using pre-trained Word2Vec embeddings, they reduce the need for large datasets, providing a fast and accurate model for text classification.
Locality Sensitive Hashing (LHS) is a common algorithm to approximate a near neighbor similarity in high dimensional spaces. Using this algorithm with Apache Spark and TensorFlow, Pinterest’s Andrey Gusev in his talk, Image Similarity Detection at Scale Using LSH and TensorFlow, will detail their system how they can detect similarity during their search over billion of items on daily basis.
Now, if you love Tulips and you have been to Holland for this flowery festival, you probably marvel at how Royal Flora Holland automate their auction processes at scale: 100K transactions per day and 400K different types of flowers and plants for their marketplace. No surprise that at the heart of their automated visual detection and operation are Keras and TensorFlow frameworks using deep neural networks and transfer learning. In this fascinating session, Operation Tulip: Using Deep Learning Models to Automate Auction Processes, Rodrigo Agundez will reveal the technical mastery.
Last year at the then Spark Summit, Messrs Andy Feng and Lee Yang introduced TensorFlowOnSpark, leveraging distributed TensorFlow training and inference by using PySpark’s distributed capabilities. Since then this framework has gained traction within the Spark community. This time around, in their talk, TensorFlowOnSpark Enhanced: Scala, Pipelines, and Beyond, they will share their new API for Spark ML pipelines to train TensorFlow models, along with support for Keras API and TensorFlow datasets.
At Databricks we cherish our founders’ academic roots, so all previous summits have had research tracks. Research in technology heralds paradigm shifts—for instance, at UC Berkeley AMPLab, it led to Spark; at Google, it led to TensorFlow. Two academics, Tiark Rompf and Gregory Essertel, from Purdue University, will share their research work on Flare and TensorFlare: Native Compilation for Apache Spark and TensorFlow Pipelines.
And finally, if you’re new to TensorFlow or Keras and want to learn how it all fits in the grand scheme of data and AI, you can enroll in a training course offered both on AWS and Azure: Understand and Apply Deep Learning with Keras, TensorFlow, and Apache Spark. Or to get a cursory and curated Tale of Three Deep Learning Frameworks, attend Brooke Wenig’s and my talk. What’s more, if you want to discern hype from reality about deep learning, check out Sameer Farooqui’s deep-dive session on Separating Hype from Reality in Deep Learning.
What's Next
You can also peruse and pick from the schedule, too. In the next blog, I will share my picks from related AI use cases, Data Science, and Productionizing Machine Learning sessions.
Take advantage of the promo code JulesPicks for a $300 discount and register now!!
Read More
Find out 5 Reasons to Attend Spark + AI Summit.