Artificial intelligence (AI) is transforming every industry. Data science and machine learning are opening new doors in process automation, predictive analytics, and decision optimization. This track offers sessions spanning the entire data science lifecycle: development, test, and production.
You’ll see examples of innovative analytics applications and systems for data visualization, statistics, machine learning, cognitive systems, and deep learning. We’ll show you how to use modern open source workbenches to develop, test, and evaluate advanced AI models before deploying them. You’ll hear from leading researchers, data scientists, analysts, and practitioners who are driving innovation in AI and data science.
Sample technologies: TensorFlow, Keras, Apache Spark, PyTorch, Apache MXNet, Theano, DL4J, R, scikit-learn, DSX, Apache Zeppelin
A modern data architecture enables enterprises to scale along with their data growth, provides flexibility to consume any and all data sources, and provides platforms to drive deep insights from the latest open source analytical tools. Striking the right balance between data strategy and cloud strategy is the first step. For many enterprises a hybrid multi-cloud data architecture that optimizes their information architecture between on-premises and the cloud is critical. It also needs to provide a global and integrated view of all their data with consistent operations, governance, and security.
This track provides the latest best practices on how to build modern data architectures. You’ll learn about key open source projects, including Apache Hadoop and related technologies, and how they integrate with the latest cloud offering to enable these transformative changes. You’ll interact with technical leads, committers, and experts who are driving the roadmaps, key features, and research around what is coming next and the extended open source big data architecture.
Data engineers and architects use multiple engines to process data in the most appropriate way, from batch ETL, to interactive SQL, to low latency NoSQL. Sessions will cover the SQL engines and tools that help users to derive the most from their data on premises and in the cloud and enrich their enterprise data warehouse (EDW).
You’ll learn how NoSQL stores like Apache HBase are adding transactional capabilities that bring traditional operational data store (ODS) workloads to Hadoop and why data preparation is a key workload. You’ll meet Apache community rock stars and learn how these innovators are building the applications of the future.
Sample technologies: Apache Hive, Apache Tez, Apache ORC, Apache Druid, Apache HBase, Apache Phoenix
The rapid proliferation of sensors and connected devices is fueling an explosion of data. Streaming data allows algorithms to dynamically adapt to new patterns in data, which is critical in applications like fraud detection and stock price prediction. Deploying real-time machine learning models in data streams enables insights and interactions not previously possible.
In this track you’ll learn how to apply machine learning to capture perishable insights from streaming data sources and how to interface with devices at the “jagged edge.” Sessions present new strategies and best practices for real-time data ingestion and analysis. Presenters will show how to use these technologies to develop IoT solutions and how to combine historical with streaming data to build dynamic, real-time predictive systems for actionable insights.
Sample technologies: Apache Nifi, Apache Storm, Streams Messaging Manager, Streaming Analytics Manager, Apache Flink, Apache Spark Streaming, Apache Beam, Apache Pulsar and Apache Kafka