1 paper · 1 filter
Zachary Yetman Van Egmond, Luis Rodrigues
This paper provides a least squares formulation for the training of a 2-layer convolutional neural network using quadratic activation functions, a 2-norm loss function, and no regu…