The 2nd Annual WiDS Datathon is now open and submissions are due by February 27th. (The sign up / merger deadline is February 21st!)
The datathon is a predictive analytics challenge focused on social impact, that gives participants an opportunity to work on a computer vision problem and develop a model that will help solve an important eco-challenge. Machine learning enthusiasts will get a chance to learn new skills and experienced ML practitioners will get a chance to sharpen their existing skills, while solving a challenging and socially relevant problem.
Join this webinar from 12-1 PM Pacific Time, Feb. 14th. The webinar is targeted towards beginners and novice participants. The presenters Vani Mandava and Jennifer Marsman will introduce machine learning concepts, give an overview of the WiDS Datathon on Kaggle, and provide tutorial-style guidance for one approach to solve the problem using Microsoft's custom vision service through the Python SDK.
Sign up for the #WiDSDatathon Community Mailing List https://mailman.stanford.edu/mailman/listinfo/widsdatathon
and join the conversation with #WiDSDatathon to get the latest updates!
UPDATE: Slides: http://bit.ly/WiDSdatathon2019webinarslides
UPDATE: Video recording: http://bit.ly/WiDSdatathon2019webinar_recording
The sample code for the Custom Vision service python tutorial that Vani shared in the webinar (using 200 images for training for a 0.94 score) is also posted at: https://www.kaggle.com/c/widsdatathon2019/discussion/80329