BrumAI Jan 2019 - Ping Pong, Neural Networks and Ethics!

This month we've got a fascinating talk on using machine learning to improve the analysis of fossil age during oil prospecting, we'll have a look at genetic algorithms, and then finish up with a group discussion on the House of Lord’s AI Ethics report.

Not only that, we're also being hosted by the awesome ping pong bar "Serve" in central Birmingham, so bring your A game :)

Talk 1: “Fossils in oil wells – an ensemble approach to predict age” - Guy Harrington

The age-determination of rocks in many oil wells is constrained by the fossils that the rocks contain. But the oil industry has been slow to use machine learning techniques to improve insight into oil prospects. I’m using the open-access dataset recently released by Equinor for the Volve field (offshore Norway) to wrestle with these sparse, messy and highly class-imbalanced data types to try and find workable protocols and models for predicting rock age. I will be showing images of fossils!

The talk will be given by Guy Harrington, who is a consultant geologist for PetroStrat Ltd and prior to this worked in academic research in the UK, Ireland and the USA. He is an honorary Senior Lecturer at the University of Birmingham.

Talk 2: How to copy nature with mathematics, how genetic algorithms work - Ammar Tarajia

Genetic algorithms are some of the easiest to program and simplest to train forms of machine learning, requiring very little information besides a basic structure, some input and objective function to learn what to do. Despite this, one can use these algorithms to learn to everything from walking to talking and everything between. I'll be going over how to build a basic neural network mathematically and then some approaches to train it using genetic evolutionary algorithms.

Ammar is currently a student, beginning the Computer Science tripos at Jesus College, Cambridge this year. I've worked as a freelance programmer throughout my time at school, programming games and modifications to existing applications during my free time.

Talk 3: Discussion on the Lord's AI Ethics report - led by Graham Lee.
We'd love you to join in, so if you have time, please read the report, which is available here.
https://publications.parliament.uk/pa/ld201719/ldselect/ldai/100/100.pdf

Graham is a software engineer, architect and trainer with degrees in physics and software engineering from Oxford University. His engineering background includes ARM and Facebook and he runs the Labrary, bringing software research and practice together by pairing research with experiment. https://labrary.online

As usual, there'll be free pizza & drinks, and some of the most interesting people in the Midlands! See you there!

Thanks to Wealth Wizards for providing the refreshments.
Thanks to Serve for being awesome and supporting us with a sweet venue!

Wealth Wizards: The UK's leading RoboAdviser. Helping people get control of their finances through the application of AI, Algorithms and regulated financial planning