We propose a new type of hidden layer for a multilayer perceptron, and demonstrate that it obtains the best reported performance for an MLP on the MNIST dataset.
Bounds for Vector-Valued Function Estimation
Correspondence of NNGP Kernel and the Matern Kernel
Exact Solutions of a Deep Linear Network
![QR code for arXiv:1301.5088 [stat.ML]](http://oss.arxitics.com/images/favicon.svg)