Applied Machine Learning Rhine-Main #1

Welcome to the inaugural edition of the Applied Machine Learning, Rhine-Main meetup! Polish your models, dust off your Neural Networks and more importantly, encircle your calendars!

To kick things off, we will have two talks presented by our community members. The first talk will be by Doug Sillers. He will be talking about how to identify and remove unwanted features from images. The second talk will be by us and we will show how to cluster Time Series data.

Full Agenda:
19:00 - 19:50 - Beer & Pizza
19:45 - 20:00 - Opening
20:00 - 20:30 - Spotlight Talk # 1: "Using AI to Power Your Images" by Doug Sillers
20:30 - 21:00 - Spotlight Talk # 2: "Time Series Clustering and Applications" by Patrick Helmig and Dipanshu Gupta.
21:00 - 22:00 - Open Discussion, Get Together & Networking

Spotlight Talk # 1: Using AI to Power Your Images
Presented by Doug Sillers

Doug is is a freelance mobile performance expert, having helped thousands of developers speed up their mobile apps and websites. A Google Developer Expert and the author of O’Reilly’s “High Performance Android Apps,” Doug regularly speaks at conferences, and blogs at dougsillars.com. He is currently working as a freelance digital nomad, traveling with his family in Europe.

Optimizing image content for every browser and device size can be difficult. Random cropping of images can lead to losing context and features (sometimes as extreme as lopping off the heads of your subjects). In this talk it will be presented how AI can be used to identify the important content in each image, allowing for smart cropping of images. This allows to optimize images further than perviously possible, shrinking page size, and speeding up page load times.

Spotlight Talk # 2: Time Series Clustering and Applications
Presented by Patrick Helmig and Dipanshu Gupta

Patrick Helmig is the co-founder of Rocketloop, a consulting and design service for the planning and implementation of intelligent software products. Dipanshu Gupta is a PhD student in Physics. In their talk, they will go through the steps on how to cluster time series, the best strategies to use, and methods to evaluate the clustering.

As promised, we will be sharing the files for both these talks with code for review and further work.

We are always looking for speakers, so if any of you are interested, feel free to contact the organizers!