Here are “companion web pages” to some research seminars that I have recently presented. Each web page below contains slides, code, links to papers, and a bullet point abstract.
Like most academics, I have several slide decks that I re-use for different audiences, with slight modifications for each version. If I posted each modified version as a separate entry, it would clutter this list. To avoid this, I only keep the most recent version up on this page, so the slides you see here may be slightly different from the ones I presented.
Deep Learning, Language, and Code: From Methodology to Applications and Back. A medley of deep learning for software engineering with some deep generative models.
Learning for Programs: Connecting code, statistics, semantics, and language. The “second wave” of research on deep learning for code is examining how to connect ML-based code analysis with natural langauge and program semantics.
Adventures in Neurosymbolic Learning. Using logical expressions and functional programs to define the structure of deep networks. Also some work on debugging deep networks.
August 2017: Learning Continuous Semantic Representations of Symbolic Expressions. International Conference on Machine Learning (ICML). Sydney, Australia.
June 2017: Machine Learning for Data Exploration and Generation. EdIntelligence student seminar, Edinburgh.
June 2017: Statistical Analysis of Computer Program Text sourced tech talk, Moscow, Russia.
April 2017: Statistical Analysis of Computer Program Text Amazon Development Centre Scotland.
January 2017: Machine Learning for Data Exploration, University of Glasgow.
December 2016: Learning Program Representations: Symbols to Vectors to Semantics NIPS Neural Abstract Machines & Program Induction (NAMPI) workshop. Barcelona, Spain. 10 December 2016
You and Your Code: Principles of Research Software Engineering: Lecture aimed at undergraduate computer science students to understand some of the ideas behind programming in research
This collection is a recent experiment on my part, inspired by a famous essay from Edward Tufte. Find out more about the philosophy behind the experiment.