AWS Machine Learning Blog

Your Guide to AI and Machine Learning at re:Invent 2018

re:Invent 2018 is almost here! As you plan your agenda, artificial intelligence (AI) is undoubtedly a hot topic on your list. This year we have a lot of great technical content on AI, machine learning (ML), and deep learning (DL)—with over 200 breakout sessions, hands-on workshops, deep-dive chalk talks, and more. You’ll hear success stories about machine learning on AWS firsthand from customers and partners such as Sony, Moody’s, NFL, Intuit, 21st Century Fox, Toyota, and more. This year’s re:Invent also includes the AI Summit, where thought leaders in the academic community will share their perspectives on the future of AI.

Here are a few highlights of this year’s lineup from the re:Invent session catalog to help you plan your event agenda.

Getting started with AI and ML

If you’re new to AI, here are some sessions to get you started with foundational concepts in machine learning and deep learning. There are overviews and demos of the Amazon SageMaker machine learning platform, deep learning frameworks, and our AI services for vision and language. These services require no machine learning skills to get started.

Leadership session – Machine Learning (Session AIM202)
Amazon has a long history in AI, from personalization and recommendation engines to robotics in fulfillment centers. Amazon Go, Amazon Alexa, and Amazon Prime Air are also examples. In this session, learn more about the latest machine learning services from AWS, and hear from customers who are partnering with AWS for innovative AI.

Build, Train, and Deploy ML Models Quickly and Easily with Amazon SageMaker (Session AIM404)
Amazon SageMaker is a fully-managed platform that enables developers and data scientists to quickly and easily build, train, and deploy machine learning models at any scale. In this session, learn how to use the Amazon SageMaker pre-built algorithms in key use cases, such as financial forecasting and predicting outcomes in healthcare.

Deep Learning for Developers: An Introduction (Session AIM301)
In this session, understand how deep learning works and learn key concepts such as neural networks, activation functions, and optimizers. We will show you how deep learning models improve through complex pattern recognition in pictures, text, sounds, and other data to produce more accurate insights and predictions. We will also share examples of common deep learning use cases, such as computer vision and recommendation models. Finally, we will help you understand how to get started using popular deep learning frameworks, such as TensorFlow, Apache MXNet, and PyTorch.

Get Started with Deep Learning and Computer Vision using AWS DeepLens (Workshop AIM316)
If you are new to deep learning, this workshop is for you. Learn how to build and deploy computer vision models using the AWS DeepLens deep learning-enabled video camera. Also, learn to build a machine learning application and a model from scratch using Amazon SageMaker.

Create Smart and Interactive Apps with Intelligent Language Services on AWS (Session AIM303)
Amazon brings natural language processing, automatic speech recognition, text-to-speech services, and neural machine translation technologies within reach of every developer. In this session, learn how to add intelligence to any application with machine learning services that provide language and chatbot functions. See how others are defining and building the next generation of apps that can hear, speak, understand, and interact with the world around us.

ML for common use cases

Whether you’re building applications to improve customer service or tackle complex problems like autonomous driving, these machine learning sessions and workshops will give you practical guidance on how to get started. Learn how to customize machine learning for business use cases as well as industry-specific applications for healthcare, finance, media and entertainment, automotive, and more.

Machine Learning for the Enterprise (Session AIM201)
Leading companies are using ML to power innovation across industries, including healthcare, automotive, and finance. In this session, learn how to build scalable ML solutions using the Amazon SageMaker platform, as well as our services for computer vision, language, and analytics. We will also demonstrate real-world use cases for enterprises to get more value from their data and to integrate and manage intelligent systems and processes.

Machine Learning at the Edge (Session AIM302, Workshop AIM315)
Video-based tools have enabled advancements in computer vision, such as in-vehicle use cases for AI. However, it is not always possible to send this data to the cloud to be processed. In this session, learn how to train machine learning models using Amazon SageMaker and deploy them to an edge device using AWS Greengrass, enabling you to process data quickly at the edge, even when there is no connectivity.

Unlock the Full Potential of Your Media Assets (Session AIM406)
Machine learning enables developers to build scalable solutions that maximize the use of media assets through automatic metadata extraction. From automatic transcription and language translation to face detection and celebrity recognition, ML enables you to automate manual workflows and optimize the use of your video content. In this session, learn how to use services such as Amazon Rekognition, Amazon Translate, and Amazon Comprehend to build a searchable video library, automate the creation of highlight reels, and more.

Transform the Modern “Contact Center” using Machine Learning & Analytics (Session AIM304)
Analyzing customer service interactions across channels provides a complete 360-degree view of customers. By capturing all interactions, you can better identify the root cause of issues and improve first-call resolution and customer satisfaction. In this session, learn how to use machine learning to quickly process and analyze thousands of customer conversations to gain valuable insights. With speech and text analytics, you can notice emerging service-related trends before they are escalated or identify and address a potential widespread problem at its inception.

Capture Voice of Customer Insights with NLP & Analytics (Workshop AIM415)
Understanding your customers is easier today than ever before. Natural language capabilities can capture a wealth of information, such as user sentiment and conversational intent. In this workshop, learn how to build an analytics pipeline that includes natural language processing (NLP). Learn how to use AWS services, including Amazon Comprehend and Amazon Transcribe, to process and perform analysis on customer data, such as contact center call recordings, to improve the customer experience.

Predicting the Next Oil Field in Seconds with Machine Learning (Session OIG302)
What if you could enhance seismic data with nothing but a few well-logged datasets using machine learning to automatically predict porosity, permeability, density, and any other lithology data in real time. In this session, we demonstrate how Amazon SageMaker can automate tasks to extract deeper insights to power better decisions and to reduce interpretation time from months to days. We also engage in a discussion about some of the applications for ML in oil and gas, including improving safety outcomes, improving asset management and maintenance, and optimizing well placement.

Fraud Detection and Prevention using Amazon SageMaker and Amazon Neptune (Session AIM422)
Business fraud is a growing concern across online and offline transactions. In this chalk talk, you will dive into detecting fraud using machine learning with Amazon SageMaker and Amazon Neptune. Topics include building models, such as class imbalance, as well as the different costs of false positives and false negatives. Additionally, we will cover algorithms like Linear Learners that can be used to build healthy models in such scenarios.

Accelerating Scientific Pace in the Life Sciences (Session LFS201)
In this session, gain insight into the key cloud computing trends for biotechnology, pharmaceutical, and medical device companies. Join AWS Life Sciences Business Development leads and our life sciences customers to hear about how you can use the AWS Cloud to modernize your clinical trials with new patient recruitment platforms. You will learn how to build collaborative research platforms between large pharmaceutical companies and research organizations, as well as how to leverage next-generation business intelligence to build real-world evidence platforms in the cloud using machine learning services such as Amazon SageMaker.

Advanced ML topics

For a deep dive into ML topics, we have a range of code-level sessions and workshops. Learn about Amazon SageMaker, deep learning frameworks such as TensorFlow and PyTorch, as well as how to customize and integrate machine learning services.

Build Deep Learning Applications Using TensorFlow (Session AIM401, Workshop AIM429)
The TensorFlow deep learning framework is used for developing diverse AI applications including computer vision, natural language, speech, and translation. In this session, learn how to use TensorFlow within the Amazon SageMaker machine learning platform. This code-level session will also include tutorials and examples using TensorFlow.

Build Deep Learning Applications using PyTorch (Session AIM402)
Deep learning can be used for diverse AI use cases, including computer vision and natural language. In this session, learn how to build and train convolutional neural networks using PyTorch within the Amazon SageMaker machine learning platform. This code-level session will include tutorials and examples using PyTorch.

Build Deep Learning Applications Using Apache MXNet (Session AIM407, Workshop AIM418)
The Apache MXNet deep learning framework is used for developing, training, and deploying diverse AI applications at scale, including computer vision, speech recognition, natural language processing, and more. In this session, learn how to get started with Apache MXNet on the Amazon SageMaker machine learning platform. This code-level session includes tutorials and examples using MXNet.

Integrate Amazon SageMaker with Apache Spark (Session AIM403, Workshop ANT318 )
Amazon SageMaker, our fully managed machine learning platform, comes with pre-built algorithms and popular deep learning frameworks, as well as an Apache Spark library that you can use to easily train models from your Spark clusters. In this code-level session, learn how to integrate your Apache Spark application with Amazon SageMaker. We will also dive deep into starting training jobs from Spark, integrating training jobs in Spark pipelines, and more.

Unsupervised Learning with Amazon SageMaker (Session AIM333)
How do you use machine learning with data that is not labeled? The unsupervised learning capabilities of Amazon SageMaker can easily handle unstructured data. In this chalk talk, we will discuss the intricacies of using unsupervised learning algorithms that are built into Amazon SageMaker, including clustering with k-means and anomaly detection with Random Cut Forest.

Deep Dive on Amazon Rekognition (Session AIM307)
Join us for a deep dive on the latest features of Amazon Rekognition. Learn how to easily add intelligent image and video analysis to applications in order to automate manual workflows, enhance creativity, and provide more personalized customer experiences. We share best practices for fine-tuning and optimizing Amazon Rekognition for a variety of use cases, including moderating content, creating searchable content libraries, and integrating secondary authentication into existing applications.

Extract More Value from Video using Machine Learning (Workshop AIM409)
Video has become an increasingly successful medium for advertising, marketing, and engaging customers. However, many companies underutilize their substantial video assets because they are poorly indexed and cataloged. In this workshop, learn how to use machine learning services to gain more value from video with sentiment analysis, facial detection and recognition, automatic transcription and translation, and more.

Better Analytics Through Natural Language Processing (Session AIM405)
Natural language processing holds the key to unlocking business value from unstructured data. Organizations that implement effective data analysis methods gain a competitive advantage through improved decision-making, risk reduction, or enhanced customer experience. In this session, learn how to easily process, analyze, and visualize data by pairing Amazon Comprehend with services like Amazon Relational Database Service (Amazon RDS), Amazon Elasticsearch Service, and Amazon Neptune. We also share real-world examples of how customers built text analytics solutions with Amazon Comprehend.

Industrialize Machine Learning Using CI/CD Techniques (Session FSV304)
As financial institutions look to accelerate and scale their use of machine learning, they need to address questions related to specific results, such as the version of the code and the data leading to a particular inference. The use of disparate and increasingly non-traditional data sources for activities such as targeted marketing, fraud detection, and improved returns is driving a need for structured development of machine learning models. In this session, we will discuss how we can use a combination of AWS services including Amazon SageMaker, AWS CodeCommit, AWS CodeBuild, and AWS CodePipeline to create a workflow that will help financial institutions to meet their requirements and drive business results.

Distributed Deep Learning & High-Res Driving Data: Advancing Autonomous Vehicles (Session CMP304)
A key barrier to the wider adoption of deep neural networks on industrial-sized datasets is the time and resources required to train them. Data scientists and machine learning engineers continue to demand shorter training times because it provides them with increased agility to try new algorithms and optimization techniques. Come and learn how Toyota Research Institute is able to optimize AWS infrastructure to minimize deep learning training times by using distributed/multi-node training.

AI Summit

With the future of AI in mind, the AI Summit at re:Invent showcases the latest in research and emerging trends. In 30-minute Lightning Talks, leaders in the research community will share their perspectives on everything from AI-fueled cancer research to quantum computing. The AI Summit will be held on Tuesday, November 27th, from 1:00pm to 5:30pm at the Venetian Theater. Visit the session catalog to register for the event.

The Future of Mixed-Autonomy Traffic (Session AIS301)
Alexandre Bayen, UC Berkeley

How will self-driving cars change urban mobility patterns? This talk will examine scientific contributions in the field of reinforcement learning, presented in the context of enabling mixed-autonomy mobility—the gradual and complex integration of autonomous vehicles into existing traffic systems. The potential impact of a small fraction of autonomous vehicles on low-level traffic flow dynamics will be explored, using novel techniques in model-free deep reinforcement learning. Examples will be presented in the context of a new open-source computational platform and state-of-the-art microsimulation tools with deep-reinforcement libraries.

Intelligent Systems for Cancer Genomics (Session AIS301)
Mona Singh, Princeton University 

One of the most exciting frontiers in science is to build automated systems that use existing biomedical data to understand and ultimately treat human disease. The key difficulty in the case of cancer is that it is a highly heterogeneous disease, making it challenging to uncover which molecular alterations in tumors are important for disease and to predict how an individual will respond to treatment. This talk will present an overview of integrative computational methods for analyzing cancer genomes that leverage a diverse range of complementary data in order to extract biomedically relevant insights.

Delivering on the Promise of AI Together (Session AIS301)
Rohit Prasad, Amazon AI

We are living in a golden age of artificial intelligence (AI). Machines have already surpassed humans in some specific tasks, including image and speech recognition, thanks to the power of cloud computing, the abundance of data required to train AI systems, and improvements in foundational AI algorithms. While some express fear about the potential for AI systems to increasingly overtake the role of humans, together we should influence how these systems can improve every aspect of our lives. Join Rohit Prasad as he explores the opportunities for AI systems to augment human intelligence in ways that will make it accessible to everyone for societal good today and into the future.

Pragmatic Quantum Machine Learning Today (Session AIS301)
Peter Wittek, University of Toronto

Quantum computing’s theoretical potential to exponentially speed up deep learning stands in sharp contrast to the current reality. Implementations are imperfect, suffering from noise and poor coherence times, and scalability limitations. In this talk, we’ll explore how quantum-enhanced machine learning plays a complementary role to classical techniques, rather than acting as a replacement. We’ll discuss relevant computing paradigms such as quantum annealing and gate-model quantum computing over discrete or continuous variables that are performed efficiently with hybrid classical-quantum protocols.

Unbiased Learning from Biased User Feedback (Session AIS301)
Thorsten Joachims, Cornell University

Logged user interactions are one of the most ubiquitous forms of data available, as they can be recorded from a variety of systems (e.g. search engines, recommender systems, ad placement) at little cost. Naively using this data, however, is prone to failure. A key problem lies in biases that systems inject into the logs by influencing where we will receive feedback (e.g. more clicks at the top of the search ranking). This talk will explore how counterfactual inference techniques can make learning algorithms robust against bias. This makes log data accessible to a broad range of learning algorithms, from Ranking SVMs to Deep Networks.

Designing for a Data-Driven Economy (Session AIS301)
Jodi Forlizzi, Carnegie Mellon University

The abundance of data available today has been described as a sea change and its own economy. Data has enabled new products, services, businesses, and economies. How can designers thrive as data-savvy innovators in this new economy? What do designers need to know about data, machine learning, and artificial intelligence? In this talk, Jodi Forlizzi will draw from multiple research and development efforts to present relevant findings about how to design in a new data-driven economy.

Making Aerial Micro-Drones a Reality (Session AIS301)
Shyam Gollakota, University of Washington

The concept of insect-scale aerial drones has long been in the realm of science fiction rather than reality, especially since powering such small drones is fundamentally difficult. In this talk, Shyam Gollakota will share his work on RoboFly, the first honeybee-sized wireless drone to successfully lift off. He will also discuss an alternative biology-based solution that integrates sensing, computing, and communication functions onto live-flying bumblebees. Data generated from this mobile IoT platform can feed AI models that have the potential to generate valuable intelligence for applications ranging from precision irrigation to environmental sensing.



About the author

Cynthya Peranandam is a Principal Marketing Manager for AWS artificial intelligence solutions, helping customers use deep learning to provide business value. In her spare time she likes to run and listen to music.