Talk Abstract: Demystifying Deep Learning for Decision Makers. One of the most important skills for data scientists to have is being able to clearly communicate results so different stakeholders and decision-makers can understand. Since data projects are collaborative across functions and data science results are…
Talk Abstract: Building an emotion detection engine (EmotionXD) and how it’s shaping what brands do.  EmotionXD – a product we developed at Digitas, is able to ingest large volumes of text, process it for topics and determine the emotion balance across eight different emotions. The…
Talk Abstract: Building an emotion detection engine (EmotionXD) and how it’s shaping what brands do.  EmotionXD – a product we developed at Digitas, is able to ingest large volumes of text, process it for topics and determine the emotion balance across eight different emotions. The…
Talk Abstract: Estimating Average Treatment Effects in Cluster-Randomised Experiments. In many experiments, the unit of randomisation is not equal to the unit of analysis. A simple example is an A/B test where users are randomly assigned to either treatment or control, but the metric of…
Talk Abstract: David is the founder of Data Idols and the Data Science Festival. He has spent years as a contract database administrator, delivering projects for some of the UK’s leading companies.  During his time as a contractor, much of his work was sourced through…
Skyscanner
VIEW BIO
Talk Abstract: TBC Bio: Data Scientist at Skyscanner
Publicis Sapient
VIEW BIO
Talk Abstract: How to make ML model fair and discrimination-free. The suitability of Machine Learning models is traditionally measured on its accuracy. A highly accurate model based on accuracy metrics are considered to be high performing models. Rarely, a data scientist tries to dissect a…
Talk Abstract: Data-driven recommendations in a public service organisation. The BBC is one of the world’s leading broadcasters, producing a large amount of content for different audiences. Data-driven recommendations are a successful approach to increase user engagement providing tailored content and personalising their experience. However,…
Royal Mail
VIEW BIO
Talk Abstract: Going postal: What we’ve learnt from trying to in-house a strategically & scientifically challenging optimisation project. Royal Mail is the UKs national postal service, delivering 16 billion items to 50 million addresses annually. We’ve been delivering post for over 500 years and we…
Talk Abstract: Personalising Depop’s browse experience. Every month millions of Depop’s users come to the platform in hope of finding that little something to complete their amazing style. To enable their creativity and simplify their experience, we have created our browse experience. In this talk,…
Talk Abstract: How does Revolut estimate the customer lifetime value (CLV) for 10M+ users across 20+ countries? At Revolut, customers come first. Building a thorough customer relationship strategy is of paramount importance. To do so, our business leaders are making data-driven decisions regarding not only…
Advancing Analytics
VIEW BIO
Talk Abstract: Scala for Big Data the Big Picture. An opportunity to explore Scala, and why it is truly a “Data Engineers language”. Scala can be a daunting language to learn, especially if you're want to take full advantage of the Functional Programming paradigm. However,…
Talk Abstract: Data-Driven Defenses and Using Data for Good.  Bio: Veronika Belokhvostova is a Director of Data Science at Facebook. In this role she oversees Integrity and Social Impact teams in the US and Europe. Integrity teams focus on protecting 2+B Facebook users from hate…
Talk Abstract: Causality-Infused Automated Machine Learning. The world is becoming more dynamic and real-time. We see more and more services becoming on-demand including food, transportation, delivery etc. This trend will continue to grow and become dominant. For example, there are more than 5 million drivers…
Talk Abstract: How does Revolut estimate the customer lifetime value (CLV) for 10M+ users across 20+ countries? At Revolut, customers come first. Building a thorough customer relationship strategy is of paramount importance. To do so, our business leaders are making data-driven decisions regarding not only…
Talk Abstract: Serve billions of real-time dynamically scaling predictions in Python: a DevOps and Data Science venture. At Unruly, we serve thousands of predictions per second with an HTTP API, with a maximum latency of less than 20ms, using an infrastructure we develop in the…
Talk Abstract: Going postal: What we’ve learnt from trying to in-house a strategically & scientifically challenging optimisation project. Royal Mail is the UKs national postal service, delivering 16 billion items to 50 million addresses annually. We’ve been delivering post for over 500 years and we…
Talk Abstract: Content is King: video - the latest frontier of Big Data. In a world where video content is being produced in every growing volumes, explore the opportunities and challenges for getting data from video. Consider the different strands of data from visual to…
Talk Abstract: The Hitchhiker's Guide to Natural Language Processing. I have developed a deep learning text matching system using neural nets with word vector embeddings, class sensitive cost functions and 1D convolution layers, to match adverts in multiple languages to vehicles on the company database.…
Applied Data Science partners
VIEW BIO
Talk Abstract: Generative Deep Learning - The Key to Unlocking Artificial General Intelligence? The talk will cover: 1. A general introduction to Generative Modelling 2. A walkthrough of one of the most utilised generative deep learning models - the Variational Autoencoder (VAE) 3. Examples of…
Talk Abstract: Rage Against the Machine: Fighting Bias in Data Science. Companies around the world are becoming more data-driven in order to be more objective, and less reliant on gut instinct and human bias. But what happens when the data or the models themselves become…
DataKind UK & The Wellcome Trust
VIEW BIO
Talk Abstract: Doing data science for good - learnings from the frontline. DataKind UK is a charity that helps social change organisations use data science, through volunteer-led projects. Today we will give you a taste of some of things we've learnt from working with charities…