Inductive Bias of Multi-Channel Linear Convolutional Networks with Bounded Weight Norm
share
We study the function space characterization of the inductive bias resulting from controlling the ℓ_2 norm of the weights in linear convolutional networks. We view this in terms of an induced regularizer in the function space given by the minimum norm of weights required to realize a linear function. For two layer linear convolutional networks with C output channels and kernel size K, we show the following: (a) If the inputs to the network have a single channel, the induced regularizer for any K is a norm given by a semidefinite program (SDP) that is independent of the number of output channels C. We further validate these results through a binary classification task on MNIST. (b) In contrast, for networks with multi-channel inputs, multiple output channels can be necessary to merely realize all matrix-valued linear functions and thus the inductive bias does depend on C. Further, for sufficiently large C, the induced regularizer for K=1 and K=D are the nuclear norm and the ℓ_2,1 group-sparse norm, respectively, of the Fourier coefficients – both of which promote sparse structures.
READ FULL TEXT