Ukrainian Data Science Club and Rails Reactor are happy to announce the next Machine Learning Training. Stay tuned for upcoming announcements.
What to expect from the upcoming training?
-> Topic 1
Kaggle: Talking Data Ad Tracking Fraud Detection Challenge - by Dmitry Danevskyi and Konstantin Maksimov
Our team of 6 (ppleskov kruegger johnpateha maksimovka yaroshevskiy ddanevskyi) has reached top 1% (8 of 3967).
The goal was to predict app downloads. Some of the downloads were known to be fraudulent.
We will tell about our approaches to feature generation, model selection and cover some of the ideas from the winning solutions.
-> Topic 2
Kaggle: 2018 Data Science Bowl - by Glib Kechyn
Glib with his team reached the 10th place and top 1% (10 of 3634) in 2018 Data Science Bowl competition.
The community was challenged to create an algorithm to automate nucleus detection (image processing task to detect and segmented nucleus).
Glib will tell about two most popular strategies for solving such kind of competition - MaskRCNN and Unet with watershed post-processing.
-> Participation is free, registration required
In order to participate please fulfill the registration form https://goo.gl/forms/WzIrEFeuM031wyq23
-> -> Who will be interested?
This training will be interesting to ML engineers and data scientists with practical experience. It might be complicated for guys without ML experience.
-> Online streaming
The training will be broadcasted on Facebook and uploaded to Facebook and YouTube. Subscribe to this event to get a link to online training.
-> Stay tuned