@misc{indiciae65b061ec6a37,
  title = {Training Neural Networks as Learning Data-adaptive Kernels: Provable Representation and Approximation Benefits},
  author = {Xialiang Dou and Tengyuan Liang},
  year = {2019},
  doi = {10.1080/01621459.2020.1745812},
  url = {https://arxiv.org/abs/1901.07114},
  note = {Source identifier: 1901.07114}
}