Explicitly Imposing Constraints in Deep Networks Via Conditional Gradients Gives Improved Generalization and Faster Convergence. Ravi, S. N., Dinh, T., Lokhande, V., & Singh, V. In AAAI, 2019.
Explicitly Imposing Constraints in Deep Networks Via Conditional Gradients Gives Improved Generalization and Faster Convergence [link]Paper  bibtex   
@inproceedings{deepcg,
  title={Explicitly Imposing Constraints in Deep Networks Via Conditional Gradients Gives Improved Generalization and Faster Convergence},
  author={Ravi, Sathya N. and Dinh, Tuan and Lokhande, Vishnu and Singh, Vikas},
  booktitle={AAAI},
  year={2019},
  URL = "https://arxiv.org/abs/1803.06453",
}

Downloads: 0