This month's Bash will focus on the much talked about subject of Machine Learning.
Dr Frances Buontempo - Regression to Mediocrity?
AI and Machine Learning appears to be the new kids on the block, but are actually as old as computers. Sometimes they are rudely referred to as curve fitting or regression. Frances (Fran) will zoom through a history of stats and machine learning and ask why it's called regression anyway.
She will encourage devs to understand what machine learning is about and data scientists to realise what we already know about development, in terms of code structure and testing. Together we can make the world a better place.
Dr Frances Buontempo wrote a book on genetic algorithms and machine learning, has given conference talks and workshops, and works as a coder, adding tests and deleting code whenever possible. You can call her Fran.
Colin Gillespie - Is ML the death of statistics?
Machine Learning has moved from dusty academic research to almost every industry. All companies now tout their cutting edge, cloud-based Machine Learning solution. Clearly, this must spell the end of Statistics. In this talk, he will discuss when machine learning is appropriate, when it's not and the bit in between. He will also discuss different real-life examples he has come across over the last few years and highlight the need for a joint approach.
Colin is the co-founder of the data science company, Jumping Rivers and is also a senior statistics lecturer at Newcastle University. Over the last few years, he has worked in statistics, machine learning and data engineering fields. He has published over seventy statistical & machine learning peer-reviewed publications and worked with companies around the world.
Dr Vicky Moulds - Engineers learning Machine Learning
Back in 2012 a Harvard Business review stated that Data Science was the sexiest job of the 21st Century. However, one of the top barriers facing companies executing on their AI / ML initiatives is a lack of skilled talent. Not only is data science and machine learning a necessary skill but so too is software engineering.
As an experienced engineer she wants to share her current journey into the machine learning and data science world. She hopes to demystify data science and machine learning concepts for engineers and walk through an example project to make it real. She think it's an exciting area and hopes to inspire other people to get involved by showing that if she can do it so can anyone.
Dani Papamaximou - The application of game theory in AI planning
Let's use AI planning to delve into corporate politics. If a team in a corporation is equated to a system and the members of staff are assumed to be agents then game theory can be applied to investigate the rate at which the performance of the team - and therefore the quality of project delivery - will degrade due to the selfish behaviour of the team members. This can be measured using the Price of Anarchy.
She will talk about the application of game theory in AI planning in decision making and the challenges arising from this approach, followed by a description of the Prisoner’s dilemma and its application in behavioural metrics. The Price of Anarchy (PoA) will then be defined followed by the presentation of the selfish routing and of different types of the PoA depending on the equilibria and the information achieved. She will also talk about prediction models and outcomes based on this approach. Finally, the potential application of the PoA approach in corporate environments will be discussed.
You can register for this event via our Bash Meetup page.