Technology hype cycles are part of the Game, but this feels different. The frenetic pace of technological change and the subsequent anxiety we feel is showing up more frequently in the stories we tell ourselves.

In the interim, you sense a Gold Rush in the making and everyone wants a cut.

These are strange reactions to a technology whose definition evades concretization.

What is it about Artificial Intelligence that inspires and terrifies us in equal parts? The 21st Century will be defined by a handful of technologies and this is one of them. So, what is it? How does it work? What is real and what is nonsense? How do we maximize the potential benefits? Can we avoid the dehumanizing aspects? Or is the whole thing just smoke and mirrors?

Brian will explore AI’s past, discuss its present and contemplate its future.

In this talk Simon will dig into the subtleties of making recommendation algorithms a seamless and integral part of your UX (goal: it should completely fade into the background. The user should not be aware they are interacting with any kind of machine learning, it should just feel right, perhaps smart or even a tad like cheating); how to solve the cold start problem (and having little training data in general); and how to effectively collect feedback data.

He'll be drawing from his experiences of building Metabase, an open source analytics/BI tool, where they extensively use recommendations and ranking to keep users in a state of flow when exploring data; to help with discoverability; and as a way to gently teach analysis and visualization best practices; all on the way towards building an AI data scientist.

In “Data: A Guide To Humans,” Phil and Noelia explore the important human skill of empathy and unpack how this skill directly affects your technical work.

In this talk they will walk you through the core topics: * Real world examples of human failures in empathy – from technology and Artificial Intelligence – to our impact on the planet, other forms of life and climate change * How empathy forms the basis of solutions – a clear, learnable model of empathy you can use and apply now.

Building this kind of stuff in the cloud is not that complicated, but if you have to make it near real-time, fault tolerant and 24/7 available, well... that's another story. In this talk, Luciano and Domagoj will tell you how they achieved this ambitious goal and how they missed a few good nights of sleep while trying to do that!

Spoiler alert: contains AWS, lambda, elastic search, monitoring, alerting & more!

Machine Learning and Deep Learning are greatly completing this picture, in a cross-platform environment, using .NET Core, Raspberry Pi 3, ML.NET, and Tensorflow. Let's deep dive into some real-world code samples, and demonstrate that C# programmers are ready for AI.

How data infiltrates the organization? Which roles come first, what problems they solve and what problems do they introduce? A down-to-earth approach in this hype driven environment to make decisions impactful and practical based on real world experience, not product brochures and GitHub repository stars.

Emotional Intelligence must complement AI for exponential growth in order to realise the true business value of analytics and machine learning.

This talk takes you on a business value perspective journey. What are the implications for technology leaders and their teams? What is the role of big data and how can we explore this?

This is due to many factors and in essence the way that the decision is being explained to the user. One potential purpose of explanations is to build the trust between the user and the AI application.

This can be achieved by providing the user with understanding of the scope of automated decision-making, and the reasons that led to a particular decision. Since, the GDPR does not appear to require accessing the “black box” to explain the internal logic of the decision-making system to data subjects. Therefore, with this in mind, this talk will try to answer the question “How can we build a trusted worthy automated decisions?”

By using the “black art power” of explanation techniques, the automated decisions can be fairer, more transparent and much understandable by the user. Further, during this talk, a framework will be presented to suggest two recommended approaches for developing the explain-ability with highlighting some of the possible challenges.

In this talk, Barbara will make the introduction to NLP concepts and deep learning architectures. The audience will be walked through two labs: sentiment analysis and text generation.

After this session, the audience will have a good understanding of the deep learning concepts when it comes to NLP. The attendees will create a classifying model that takes advantage of the document context using TensorFlow library and scale their solutions using Google Cloud.

Just as humans require contextual information to make better decisions, so do machine-learning algorithms. Combining ML processing with a graph data structure can help fill in the missing contextual information and improve our predictions.

In this session, Mark will show what graph has to offer and show an example applying link prediction analysis to estimate how likely academic authors are to collaborate with new co-authors in the future. You will see how to fine-tune the elements you measure and understand the results for decisions or further adjustments. Learn how to exploit the power of connected data to improve prediction analysis!

You will learn what are meta reinforcement learning algorithms, how they achieve State-of-the-Art performance in multiple machine learning tasks and their potential applications.in different industries.

In this talk, Samantha will look at the landscape of messy and patchy healthcare data, understand the difficulties of drawing reasonable conclusions from the data, and discuss the challenges of changing user behavior in healthcare.

Laura will explain the main challenges companies are facing in the Data Science adoption and how the use of automated ML and democratisation is helping them to success.

The CAP theorem is widely known for distributed systems, but it's not the only tradeoff you should be aware of. For datastores there is also the FAB theory and just like with the CAP theorem you can only pick two:

Fast: Results are real-time or near real-time instead of batch oriented Accurate: Answers are exact and don't have a margin of error. Big: You require horizontal scaling and need to distribute your data.

While Fast and Big are relatively easy to understand, Accurate is a bit harder to picture. This talk shows some concrete examples of accuracy tradeoffs Elasticsearch can take for terms aggregations, cardinality aggregations with HyperLogLog++, and the IDF part of full-text search. Or how to trade some speed or the distribution for more accuracy.

When big ideas meet and reinforce one another constructively, we see even larger effects. Event streaming has already caused a quiet revolution in the traditional ETL space, and is helping microservices architectures deliver on their yet-elusive promises of decoupling and evolvability. When event streaming and machine learning meet, the well-known problems of production ML systems find the same kind of solutions. Machine learning and event streaming can deliver all kinds of value to the world separately, but the ML community stands to grow by adopting real-time retraining using platforms like Apache Kafka.