Delve into the future of research at MiS with our preprint repository. Our scientists are making groundbreaking discoveries and sharing their latest findings before they are published. Explore repository to stay up-to-date on the newest developments and breakthroughs.
MiS Preprint
63/2020
Implicit bias of gradient descent for mean squared error regression with wide neural networks
Hui Jin and Guido Montúfar
Abstract
We investigate gradient descent training of wide neural networks and the corresponding implicit bias in function space. Focusing on 1D regression, we show that the solution of training a width-n shallow ReLU network is within n^{−1/2} of the function which fits the training data and whose difference from initialization has smallest 2-norm of the second derivative weighted by 1/ζ. The curvature penalty function 1/ζ is expressed in terms of the probability distribution that is utilized to initialize the network parameters, and we compute it explicitly for various common initialization procedures. For instance, asymmetric initialization with a uniform distribution yields a constant curvature penalty, and thence the solution function is the natural cubic spline interpolation of the training data. The statement generalizes to the training trajectories, which in turn are captured by trajectories of spatially adaptive smoothing splines with decreasing regularization strength.
