Reinforcement Learning, Visual Recognition, OpenShift... & the rest! A review of Tuesday night's Meetup

Tuesday night saw us return to the Bank of Ireland Workbench down on Grand Canal Square (thanks Darko!) where we covered three very interesting topics, that to be fair you could do a whole meetup on each!

First up, Brad took us through Reinforcement Learning, how that differs from Machine Learning, where a system makes decisions based on a system of rewards as it tries to balance unexplored territory vs. its current knowledge. Brad had some great examples of how this is used in gaming (Google’s AlphaGo), traffic patterns, autonomous cars and some pretty terrifying robots. In terms of robots, Brad also showed us OpenAI Gym and Gazebo, some very cool tools for RL in Robotics.

Second up, Rory was kind enough to share his Watson Visualisation project, and how it reached the finals of the 2018 Watson Build Challenge. Rory went deep on the tech, but also had loads of helpful tips, tricks and advice when it comes to image recognition, classification and anomaly detection. Throw in a bit of blockchain and some potential use cases, and there was something for everyone.

After a great networking session with the usual pizza and refreshments, Brad was up again to give us his view of OpenShift. It went from the basic (what is it, the history) right through to the architecture and why you should use it. Brad also directed us to OKD, MiniShift, a free demo, and the S2I framework.

All in all a super evening, thanks to the speakers and the attendees, some fantastic Q&A after each presentation - we'll be back on 5th September so keep an eye out on Meetup.com for the topics and location. As always we welcome your feedback, and you can find all the content from our 2019 Meetups here!