===============================================================
For citation of our work:
@inproceedings{AQM2016,
  title={Globally Optimal Training of Generalized Polynomial Neural Networks with Nonlinear Spectral Methods},
  author={A. Gautier and Q. Nguyen and M. Hein},
  booktitle={Advances in Neural Information Processing Systems (NIPS)},
  year={2016}
}

The following version presents a general theory for more general class of non-convex optimization problems:
@inproceedings{QAM2016,
  title={Nonlinear Spectral Methods for Nonconvex Optimization with Global Optimality},
  author={Q. Nguyen and A. Gautier and M. Hein},
  booktitle={NIPS Workshop on Optimization for Machine Learning},
  year={2016}
}
===================================================================
Installation:
	cvx is required: http://cvxr.com/cvx/download/
only for train_SGD.m as it requires projection onto p-norm sphere.

===================================================================
Guideline: 
Please see the following files to run our experiments  
0. NLSM_demo.m: demo our NLSM on cancer/iris dataset  
1. main_NLSM.m: testing our Nonlinear Spectral Method  
2. main_ReLU1.m: testing one-hidden-layer ReLU nets by Batch-SGD  
3. main_ReLU2.m: testing two-hidden-layer ReLU nets by Batch-SGD  

In all experiments, we use UCI-datasets obtained from:  
https://archive.ics.uci.edu/ml/datasets.html


