Elias Wirth (TU Berlin/Zuse-Institut Berlin)
Due to the current situation, the talk takes place online, via zoom. The meeting link has been sent out via the usual mailing lists; please contact the organisers if you have not received the email and would like to join the talk.
We introduce the Frank-Wolfe Algorithm, also known as Conditional Gradients.
Recently, the algorithm has been revisited for various applications in machine learning. We motivate the Frank-Wolfe Algorithm and discuss the properties that make it appealing for practitioners.