Ensemble Methods and Dialogue Systems

Welcome to the inaugural London Applied Deep Learning Meetup!

The format of the meetup is two talks: an entry level talk for those who are just starting to get into machine learning, followed by a more advance talk by researchers and engineers, in academia or industry, on current research or other topics that is of interest.

We're very grateful for Entrepreneur First to host the meetup at their headquarters in Bermondsey. Refreshments will be provided.

Plan:
-- 7:00pm Doors Open
-- 7:15pm Introduction to Ensemble Methods
-- 7:45pm Break
-- 8:00pm Dialogue Systems
-- 8:30pm Drinks

Introduction to Ensemble Methods - by Alan Mosca:
In this talk, Alan will cover the basics of Ensemble theory, and explore some options on how ensembles can be applied in modern Deep Learning. He will then examine the benefits of combining Ensembles with other techniques, and what this means both for improving the performance of models, and increasing the resilience to adversarial attacks.

Bio:
Alan is the CTO of nPlan, an EF startup that uses deep learning to change construction project management, and a part-time researcher and associate lecturer at Birkbeck, University of London. His research is in Deep Learning Ensemble methods and non-convex optimization.

Democratise Conversational AI - Scaling Academic Research to Industrial Applications - by Tsung-Hsien (Shawn) Wen:
Deep learning has had a profound impact on conversational AI research. In recent years, models such as CNN intent extractors, DQN policy networks and LSTM language generators have been at the centre of spoken dialogue systems research. Despite the transformational potential of these methods, not all of these approaches are ready for production. For various reasons, these methods struggle to scale to complex, real-world conversational scenarios. In this talk, I will share the insights we gained from building conversational agents in academia, and how these unique experiences are empowering PolyAI and allowing us to scale across multiple application domains and languages.

Bio:
Tsung-Hsien (Shawn) Wen is a co-founder and the Chief Scientist of PolyAI, a London-based startup looking to use the latest developments in NLP and ML to create a general platform for deploying spoken dialogue systems. He holds a PhD from the Dialogue Systems group, University of Cambridge, where he worked under the supervision of Professor Steve Young.
His research focuses on language generation and end-to-end dialogue modelling, specifically in learning to generate responses for task-oriented dialogue systems. His work on natural language generation received best paper awards at EMNLP 2015 and SigDial 2015. He gave the “Deep Learning and NLG” tutorial at INLG 2016 and has given invited seminars to research groups at Google, Apple, Xerox, and Baidu China, as well as Samsung’s corporate training course in Warsaw. Before starting PolyAI, Shawn was a research consultant for IPSoft Amelia, and a research intern at Google Brain.