Conditional Gradients - an introduction! In this talk Sebastian will give a brief introduction to conditional gradients methods, which are often the method of choice for many machine learning problems when it comes to sparsity (and associated statistical guarantees) and projection-freeness. Important applications include matrix completion and recommender systems, compressed sensing, sparse signal recovery, sparse regression, identification of predictive features in neural networks and many more. He will also present several such examples and applications in this talk.
