Improper Deep Kernels

@inproceedings{Heinemann2016ImproperDK,
  title={Improper Deep Kernels},
  author={Uri Heinemann and Roi Livni and Elad Eban and Gal Elidan and Amir Globerson},
  booktitle={AISTATS},
  year={2016}
}
Neural networks have recently re-emerged as a powerful hypothesis class, yielding impressive classification accuracy in multiple domains. However, their training is a non-convex optimization problem which poses theoretical and practical challenges. Here we address this difficulty by turning to “improper” learning of neural nets. In other words, we learn a… CONTINUE READING