On Technical User Groups - Part 2

To follow our post on technical user groups, here's a hat-tip to meetups & conferences throughout UK, Ireland and Europe that we enjoyed attending in 2015.

We previously detailed why technical user groups are important and the many attendant benefits for personal continuous development and community building.

This seemed to ring true with many of our readers, so what follows is a quick roll-call and thanks to some of the many user-groups that we each have attended, spoken at, or otherwise helped to build in 2015.



As ever, London has a large tech community and very rapidly growing groups in the FinTech (finance tech) and InsTech (insurance tech) areas. In particular it's been very interesting to attend:

  • Instech London and Insurance Tech with both technical briefings and startup pitches on the potential impacts upon insurance & regulation of advanced data science, blockchain databases, consumer on-demand applications, IoT (Internet of Things) and more
  • Fintech Storm and Fintech Tuesday with similar briefings by industry experts on how new technology and analysis is impacting finance, banking and law
  • Big O London, London Machine Learning and Bayesian Networks London for reliably interesting talks at the forefront of academic research and industrial application
  • PyData London and Python for Quant Finance, respectively hugely popular and very rapidly growing groups oriented toward talks on practical uses of Python in data science and quantitative finance. Jon has given talks at both recently, on using t-SNE for efficient visualisation of high-dimensional datasets.

Dublin & Belfast

Data science groups are also in rude health where the grass grows greener:

  • Dublin R continues to be a stalwart of the local statistics community and naturally Mick Cooney has given a number of talks and workshops over the year, including Linear Dynamical Systems and Probabilistic Graphical Models
  • Several new groups have sprung up over the year, covering many aspects of data science / machine learning and we wish them well, including Data Scientists Ireland, Dublin Spark and Machine Learning Dublin
  • Big Data Belfast Breakout is curated and sponsered by Analytics Engines and is well attended by local industry and academia. Mick Crawford gave a lighthearted talk at their December meetup on our experiences working on a data science project for a large insurance company. We wanted to give an idea of how the experimental nature of a data science project can lead you into the weeds, but also how that experience can be utilised, learned from and used to generated valuable insights. Occasionally you find some gems while you're in there. We demonstrated an application to augment insurance policy data with socioeconomic data derived from the Irish Census that came out of this process. This application has a multiple of applications not limited to the insurance industry.

... and Away

We've also occasionally ventured away from our usual stomping grounds, to visit, learn from and present at meetups and conferences elsewhere in Europe. Hopefully we'll continue this trend in 2016.

Köln R

Mick Cooney recently went over to speak at Köln R, here's a his quick review:

Köln R is a really cool R User Group organised by Markus Gessman and Kirill Pomogajko. They asked me to give a talk, so I went along and gave a presentation about a Bayesian approach to monitoring process change, following my earlier blogpost on the subject.

Köln and Düsseldorf are amazing cities, and the efficiency of the public transport on the continent always awes me. Sadly, it was a flying visit but I expect to be back in the near future for a more social visit. The venue was Startplatz, a tech startup incubator in the heart of the city, perfect for what we needed.

The other two talks were by Dietmar Janetzko who spoke about DataJoy, a web service that facilitates interactive teaching of R; and Jessica Peterka-Bonetta who spoke about decoding emoticons in R and explained how Paris Hilton provides a valuable social good!

Many thanks to all the interesting questions at the end - I have learned that I need to compare my approach to a simpler method using the Kolmogorov-Smirnoff statistic.

Earlier in the day, I met Thomas Wiecki, the organiser of Düsseldorf Data Science. He does a lot of work using Bayesian analysis using the Python package pymc3 and after we talked all things Bayesian, he helped me catch the train to Köln without too much drama. I'll have to go over to meet both user groups again!

R in Insurance 2015 in Amsterdam

We took a couple of days in midsummer to attend and present at the R in Insurance conference at the University of Amsterdam.

This was our second year and it was great to see a growing attendance, meet lots of likeminded practitioners and hear some really interesting talks. We wrote up some notes shortly afterwards.


It was a great year for connecting with technical and business communities old and new. We'll hope to continue the trend this year.

Finally, on that note, I'm very happy to announce we're supporting the R in Insurance conference in July 2016, this time held in London at Cass Business School. If you're at all interested in the technical aspects of insurance, data science and the future of the industry from an analytical point of view, it's well worth attending or even submitting a talk.

  1. Photo sourced from James Cridland.

Jonathan Sedar

Jon founded Applied AI in 2013. He's a specialist in Bayesian statistics and the application of machine learning to yield insights and profits throughout fintech, banking and insurance sectors.