Using Multi-Threading to Build Neural Networks with TensorFlow and Apache Spark

At the recent TensorFlow meetup in Paris, the attendees learnt how to construct neural networks employing the multi-threading approach. The topics under discussion also included distributed TensorFlow and the efficiency of using Apache Spark with TensorFlow.

TensorFlow and deep learning without a PhD

Martin Gorner of Google introduced the attendees to TensorFlow and taught how to choose the right neural network for a specific problem. He demonstrated how to train the network with just a couple of code lines and a bag of “tricks of the trade.”

Constructing neural networks using multi-threading

In his session, Maxence Queyrel of Quinten talked about multi-threading with TensorFlow and Apache Spark, highlighting:

• Neural networks hyperparameters
• How TensorFlow works with Apache Spark (examples, efficiency, etc.)
• Distributed TensorFlow: What it is, how it works, and how to use it

TensorFlow + Apache Spark

Jiqiong Qiu of SFEIR provided an overview of H2O—an open-source machine learning platform—explaining why and how to use it together with TensorFlow and Apache Spark. With H2O, users can get the best of Apache Spark (its APIs, RDD, simple context, multi-tenancy, etc.) and combine it with deep learning algorithms.

Join the meetup group to stay informed about the upcoming events.

Further reading

• Introduction to Neural Networks and Meta-Frameworks for Deep Learning with TensorFlow
• Distributed TensorFlow and Classification of Time Series Data Using Neural Networks
• Performance of Distributed TensorFlow: A Multi-Node and Multi-GPU Configuration

About the speakers

Martin Gorner is responsible for developer relations at the Paris branch of Google. He is passionate about science, technology, coding, algorithms, and everything in between. Martin has a degree in civil engineering from Mines ParisTech. He spent 11 years shaping the nascent eBook market within the Mobipocket startup, which later became part of Amazon Kindle.

Maxence Queyrel is Data Science Intern at Quinten. He works with deep learning libraries and is in charge of integrating TensorFlow into the company’s activities. Maxence is currently researching machine learning, data science, and big data as part of his master’s degree at Pierre and Marie Curie University.

Jiqiong Qiu is Data Scientist at SFEIR. She holds a master’s degree in engineering and bioinformatics, as well as a PhD in data science. Jiqiong has experience in analyzing IoT data to detect abnormal behavior of domestic animals and working on a remote diagnostic system for skin diseases. Her current job is mainly focused on training deep neural networks using TensorFlow.