Before the main PyData Global conference, the committee ran an Expert Briefings series- these were presented by individuals who are key contributors to the PyData community, and have extensive experience in their respective domains. It was in the format of a short presentation (15 mins) on the state of the art in their area of expertise, and a discussion session afterwards. Recordings of these sessions will be made available to attendees throughout the conference!
The State of the Art for Probabilistic Programming | Monday 18th 2pm UTC
How To Ensure Responsible Use Of AI With A Real-World Example | Monday 18th 6pm UTC
The State of Higher Performance Python | Wednesday 20th 4pm UTC
Natural Language Processing: Trends, Challenges and Opportunities | Wednesday 20th 5pm UTC
How You Benefit by Contributing to Open Source | Wednesday 20th 8pm UTC
Bayesian modeling is currently undergoing a Renaissance. Better and more user-friendly tools, as well as algorithms, allow this technique to be used by more people on larger and more complicated problems. While academia has already applied these powerful tools for research for a while, more and more businesses, frustrated by the empty promises of uninterpretable machine learning, are realizing the impact these more transparent methods can have.
In this session, I will give a state of the art of probabilistic programming followed by a Q&A session.
This event is for ticket holders, you’ll receive your invite by email. You are strongly encouraged to bring questions and discussion topics for the second half of this session.
Thomas Wiecki is an author of the PyMC library and founder of PyMC Labs, a Bayesian consultancy solving advanced data science problems. He did his PhD at Brown University building computational models of the brain.
The use of data and sophisticated algorithms have mushroomed over the last few years, with startups and large enterprises alike developing innovative products and services. However, concerns around safety, fairness, and ethics have also grown as we’ve seen more and more examples of failure due to bias, poor accuracy and even misapplication of algorithms. In response, governments around the world have developed standards, frameworks, and recently, proposed regulations, for assessing the quality of data-driven automation.
In this session, we will briefly review the key elements of official guidance on responsible use of data and AI.
We will then explore an example of an innovative product designed to help the more vulnerable in our communities (based on a real startup product) and, as a group, discuss how we should ensure it is developed to be safe, fair and ethical. We’ll cover not just questions of data and algorithms, but also the development process and governance. By attending this session you will gain an overview of the regulatory direction for AI, and have fun sharing your ideas and learning from others’ approaches to safe, fair and ethical AI.
This event is for ticket holders, you’ll receive your invite by email. You are strongly encouraged to bring questions and discussion topics for the second half of this session.
Over 25 years, Tariq Rashid has worked as a technologist in sectors as diverse as nuclear energy, tech startups and central government. Over that time he has observed directly how data and technology can both benefit, and harm, individuals and society.
Bringing together his experience in digital services, and his background in machine learning and data mining, he now runs a specialist consultancy providing independent assurance for data-driven automation.
Tariq is passionate about community, and runs the Data Science Cornwall and London Algorithmic Art meetups, as well as a children’s code club. He enjoys speaking at, or leading workshops, at community events and conferences.
We’ll review the state of the art in the data science world for common number-crunching tasks on small to big data. Topics we’ll cover include profiling, compilation and data manipulation. We’ll also review the near future for Numba, Pandas, Dask, Vaex, Modin, and Polars and I’ll help you make some pragmatic choices about tools you might invest time in. We’ll have plenty of time to discuss your use cases and problems you might have encountered.
Please bring questions about tools or processes that we can discuss during this session.
This event is for ticket holders, you’ll receive your invite by email. You are strongly encouraged to bring questions and discussion topics for the second half of this session.
Ian Ozsvald is a Chief Data Scientist and Coach, he’s helped co-organize the annual PyDataLondon conference with 700+ attendees and the associated 11,000+ member monthly meetup. He runs the established Mor Consulting Data Science consultancy in London, gives conference talks internationally often as keynote speaker, and is the author of the bestselling O’Reilly book High-Performance Python (2nd edition). He has 19 years of experience as a senior data science leader, trainer, and team coach. For fun, he’s walked by his high-energy Springer Spaniel, surfs the Cornish coast, and drinks fine coffee. Past talks and articles can be found at: https://ianozsvald.com/
https://github.com/ianozsvald/
Natural Language Processing (NLP) is one of the most exciting fields in AI, with successful applications ranging from personal assistants, to customer service, to healthcare and many more. In the past few years we have witnessed several breakthroughs in different NLP tasks, as well as the improvement and consolidation of the NLP open-source ecosystem. Despite the buzz, building successful and efficient NLP solutions remains a challenging endeavour.
This session starts with a short presentation that will offer an overview on the advances, challenges and opportunities in NLP technologies, looking at different modelling solutions and at the Python ecosystem. We’ll then continue with a Q&A / round table discussion. By attending this session, you’ll learn about recent trends in NLP as well as old-but-gold approaches, you’ll have the opportunity to ask your burning questions on NLP, and you’ll hopefully leave more confident and better equipped to tackle your NLP problems.
This event is for ticket holders, you’ll receive your invite by email. You are strongly encouraged to bring questions and discussion topics for the second half of this session.
Marco Bonzanini is a Data Science consultant, corporate trainer and author based in London (UK) specialised in the Python for Data Science (PyData) software stack. With 20 years of experience in the tech industry, he provides consulting, coaching and training services in the data science space through his company Bonzanini Consulting Ltd. Backed by a PhD in Information Retrieval, Marco focuses on text analytics applications and he has enjoyed working on a broad range of information management and data science projects, including flight safety, social media data, behavioural data, biomedical data and recruitment data.
He’s the author of “Mastering Social Media Mining with Python”, “Data Analysis with Python” and “Practical Python Data Science Techniques”, published by Packt Publishing. He currently serves as chair of the PyData London meetup and conference series (largest Python user group in Europe) and regularly speaks at international tech events.