Back for the fifth year running, the Deep Learning Summit will once again be joined by the Deep Learning in Healthcare Summit in Boston this May 23 - 24. In addition to these two summits, attendees will also have access to the Deep Dive Track which will provide technical insights to some of the key topics explored at the event. The two days will offer attendees the opportunity to watch presentations from global AI experts, join interactive workshops, experience interviews and fireside chats, and take part in the extensive networking sessions where attendees will all come together in the exhibition area to share their knowledge and make new connections.

With over 60 speakers from both industry and academia sharing their knowledge, all three tracks offer insights to the most cutting edge progressions in deep learning. Not sure how you’re going to spend your time at the summit? Here are some suggestions…

...we recommend the following sessions:

Session: What Do Your Neural Networks Learn? A Peek Inside the Black Box
Expert: Brandon Rohrer, Data Scientist, Facebook
Track: Deep Learning
Time: Day 1, 09:35

Deep neural networks are famously difficult to interpret. This session will take a tour of their inner workings to build an intuition of what's inside the black box and how all those cogs fit together. Then the discussion will move to use those insights as we step through an image processing problem with deep learning, showing at every step what the neural network is "thinking".

Session: Distributed Tensorflow: Scaling Model Training to Multiple GPUs
Expert: Neil Tenenholtz, Senior Data Scientist, MGH & BWH Center Clinical Data Science
Track: Deep Learning in Healthcare
Time: Day 1, 09:55

While offering state-of-the-art performance across a variety of tasks, deep learning models can be time-consuming to train, thus hindering the exploration of model architectures and hyperparameter configurations. However, this bottleneck can be greatly reduced by leveraging the near-linear speedups afforded by multi-GPU training. In this talk, Neil will explore the different manners in which Tensorflow supports training to be distributed across a collection of GPUs.

Session: Considering the Ethics in AI
Expert: Cansu Canca, AI Ethics Lab
Track: Deep Dive Workshop
Time: Day 1, 13:40

Cansu is the founder and director of the AI Ethics Lab, where she leads teams of computer scientists and legal scholars to provide ethics analysis and guidance to researchers and practitioners. She has a Ph.D. in philosophy specializing in applied ethics. She works on the ethics of technology and population-level bioethics with an interest in policy questions. Prior to the AI Ethics Lab, she was a lecturer at the University of Hong Kong, and a researcher at the Harvard Law School, Harvard School of Public Health, Harvard Medical School, Osaka University, and the World Health Organization.

Session: Deep Learning for the Future Enterprise
Expert: Gautam Shroff, VP and Chief Scientist, Tata Consultancy Services
Track: Deep Learning
Time: Day 2, 09:50

There are multiple challenges that impede the effective application of deep learning to real-world problems such as machine health monitoring, container stowage planning and healthcare recommendations. Slowly but surely, companies are finding solutions to these problems and the impact of deep learning is now percolating to enterprises in many different sectors. As one of the worlds largest IT consulting firms, TCS has nurtured a dedicated team of deep learning researchers to provide solutions to these problems for use cases across sectors ranging from manufacturing and shipping to healthcare and finance.

...we recommend the following sessions:

Session: Demystifying AI Terms & Tools - Introductory Overview to AI Key-Terms
Expert: Panel Discussion & Q&A, experts announced soon
Track: Deep Dive  
Time: Day 1, 09:55

An introduction and overview to some of the key AI terms which you can expect to hear throughout the 2-Day Summit.

Session: Natural Language Processing for Healthcare
Expert: Director of Machine Learning and AI, CODAMETRIX
Track: Deep Learning in Healthcare  
Time: Day 1, 11:40

With recent advancements in Deep Learning followed by successful deployment in natural language processing (NLP) applications such as language understanding, modeling, and translation, the general hope was to achieve yet another success in healthcare domain. Given the vast amount of healthcare data captured in Electronic Medical Records (EMR) in an unstructured fashion, there is an immediate high demand for NLP to facilitate automatic extraction and structuring of clinical data for decision support. Nevertheless, the performance of off-the-shelf NLP on healthcare data has been disappointing. Recently, tremendous efforts have been dedicated by NLP research pioneers to adapt general language NLP for healthcare domain. This talk aims to review current challenges researchers face, and furthermore, reviews some of the most recent success stories.

Session: Deep Learning for the Future Enterprise
Expert: Matthre Mattina, Senior Director of ML & AI Research, Arm Track: Deep Learning
Track: Deep Learning
Time: Day 2, 09:50

Deep neural networks are a key technology at the core of advanced audio and video applications. As these applications begin to migrate from large servers executing in the cloud to mobile and embedded platforms, they place significant demands on the underlying hardware platform. This talk will review the key properties of these models and how these properties are leveraged to deliver efficient inference on energy, compute, and space-constrained platforms.

Session: Applications of Deep Learning to New User Recommendations at Twitter
Expert: Jay Baxter, Senior Machine Learning Engineer, Twitter Cortex
Track: Deep Learning
Time: Day 2, 11:20

The cold start problem for new users is a classic challenge for recommender systems. In this talk, Jay will discuss some deep learning approaches that can be used to address this problem, including using neural networks to train co-embeddings of new users and items and serving them in an efficient way at runtime via approximate nearest neighbour algorithms like LSH or HNSW. He will also touch on some of the difficulties of evaluating such models both offline and online in the context of A/B tests.

Session: Rising Stars - Presentations from the next generation of AI Pioneers
Expert: AI4ALL innovators & AI rising stars
Track: Deep Dive
Time: Day 2, 13:50

As a CxO, you’ll be involved in the growth and hiring process of your team. In this session, you’ll learn from the next generation of experts and find out what excites them about working in the space. Learn more from young prospective leaders of the AI world as they present their latest research and find out how you could work together.

...we recommend the following sessions:

Session: Learning to Synthesize Images
Expert: Akram Bayat, Research Assistant, University of Massachusetts Boston
Track: Deep Learning
Time: Day 1, 11:20

In this talk, Akram will present deep learning solutions for three visual scene perception and object recognition problems. The goal is to investigate to which extent deep convolutional neural networks resemble the human visual system for scene perception and object recognition: (1) classification of scenes based on their global properties, (2) deploying multi-resolution technique for object recognition, and (3) evaluating the influence of the high-level context of scene grammar for object and scene recognition. The first problem proposes to drive global properties of a scene as high-level scene descriptions from deep features of convolutional neural networks in scene classification tasks. The second problem shows that fine-tuning the Faster-RCNN to multi-resolution data inspired by human multi-resolution visual system improves the network performance and robustness over a range of spatial frequencies. Finally, the third problem studies the effects of violating the high level scene syntactic and semantic rules on human eye-movement behavior and deep neural scene and object recognition networks.

Session: Predicting the Effects of Genetic Medicines Using Transfer Learning
Expert: Amit Deshwar, Director of Predictive Systems, Deep Genomics
Track: Deep Learning in Healthcare
Time: Day 1, 14:10

Genetic medicines promise the ability to precisely target the root causes of disease. At Deep Genomics, Amit and his team are developing machine learning systems to predict the properties of these medicines, including activity and safety. A fundamental problem in doing so is that large collections of therapeutic data is infeasible to collect. Using transfer learning allows them to fuse large amounts of inexpensive biology data with small amounts of therapeutic data. Amit will discuss how he has successfully used transfer learning to predict the on-target activity of genetic medicines, enabling us to test five times fewer compounds for some of their targets.

Session: Choosing Which Deep Learning Method is the Right Tool to Use - Case Studies & Best Practices
Expert: Panelists to be announced
Track: Deep Dive
Time: Day 1, 16:20

Working on AI and deep learning projects is exciting, and the desire to solve problems and make breakthroughs can often be so intense that you rush into a project before assessing your ammunition. You may have used a certain method of deep learning in a previous project which showed good results, but it’s important not to jump straight in with the same methods before assessing the problem at hand. In this session, experts will share their failures and successes and outline the best way to succeed through the careful selection of DL methods.

Session: Building Visual Search at Salesforce
Expert: Michael Sollami, Lead Data Scientist, Salesforce
Track: Deep Learning
Time: Day 2, 11:40

Learn how academic research in deep learning methods is being applied in industry. Fine-grain recognition remains an unsolved problem at in the general case, indeed, it may even be as difficult as self-driving cars. There are many technical challenges in achieving accurate production-level image retrieval at web scale. This talk will highlight the hurdles in building such a search platform. Salesforce have trained networks to discover a manifold representing the space of all consumer products. Michael will present the current architectures in embedding networks, i.e. the mapping from consumer images to the product feature space, as well as the most promising research directions. Implementation level details will be covered insofar as they make efficient fine-grain retrieval possible, and performance evaluation (both statistical as well as qualitative) measures will be described.