GDG Reading is teaming up with Women Techmakers Reading and Central Working for a belated International Women's Day event https://www.internationalwomensday.com. This is the start of more events with the theme #BalanceForBetter.
We have a great speaker line up and we would love to hear from you as well. Please fill out our CFP to submit a lightning talk or a fireside chat question: https://bit.ly/2Yky8d9
6:00 Food and networking
6:30 - 6:40 Welcome
6:40 "Talking technology" by Léonie Watson (35mins)
7:15 "Fast Adaptation via Meta Learning" by Luisa Zintgraf (25mins)
7:40 "Perfectionism, Impostor Syndrome, Anxiety and Learning to be Kind to Yourself" by Jo Franchetti (45mins)
8:25 - 8:35 Tea break
8:35 "Talk to be confirmed" by Nana Fifield
9:00 9:20 Lightning talks/fireside chat questions and discussions
9:20 Closing remarks
Léonie is Director of TetraLogical; a member of the W3C Advisory Board; co-Chair of the W3C Web Platform Working Group; and a member of the Accelerated Mobile Pages (AMP) Advisory Committee. Please read more at https://tink.uk/about-leonie.
We've been talking with technology for longer than you might think, but
despite increasing conversational sophistication, relatively few ways
exist to help us make artificial speech sound human. Find out how to use
Speech Synthesis Markup Language (SSML) with the Amazon Echo or Google Home, and the Web Speech API in the browser; how the CSS Speech module might be useful (if only it were more widely supported); and how we might be able to solve some problems by making things talk.
Luisa is a second year DPhil student, supervised by Shimon Whiteson (University of Oxford) and Katja Hofmann (Microsoft Research). Luisa has a B.Sc. in Mathematics (University of Hamburg) and a M.Sc. in Artificial Intelligence (University of Amsterdam). Before starting her PhD, she worked at the University of Brussels as a research assistant.
Her current research focuses on meta reinforcement learning, with a particular focus on how to enable artificial agents to adapt fast to new environments and tasks.
The challenge of fast adaptation in machine learning is to learn on previously unseen tasks fast and with little data. In principle, this can be achieved by leveraging knowledge obtained in other, related tasks. However, the best way to do so remains an open question. In this presentation I will talk about the meta-learning approach to fast adaptation, i.e., learning how to learn on unseen problems/datasets within a few shots. In particular, I will focus on gradient-based methods, i.e., methods that adapt to unseen tasks within just a few gradient steps. I will also talk about my current work, the algorithm CAVIA, a meta-learning method for fast adaptation that is scalable, flexible, and easy to implement. I will show some empirical results on a variety of learning problems.
Jo is an Engineering Engagement Manager at Trainline who is passionate about WebVR, PWAs, and great CSS. She’s got 7yrs of experience as a front end developer and has worked in various parts of the tech industry from startups, agencies, and charities to large organisations. She is also a mentor and organiser at codebar.io where she is able to action her passion not only for teaching good use of the web but also for improving the diversity and inclusivity of the tech industry.
Ever felt like everyone knows more than you? Spent hours procrastinating rather than starting or finishing a project? Perfectionism, anxiety & impostor syndrome affect many people in the tech industry. Learn how to spot harmful thought patterns, how to avoid anxious spirals and be kind to yourself!
We are very excited to welcome everyone to an inspirational and fun evening of learning and discussions.