Neural Architecture Search From Fréchet Task Distance

03/23/2021

∙

by   Cat P. Le, et al.

∙

Duke University

∙

Stanford University

∙

0

∙

share

We formulate a Fréchet-type asymmetric distance between tasks based on Fisher Information Matrices. We show how the distance between a target task and each task in a given set of baseline tasks can be used to reduce the neural architecture search space for the target task. The complexity reduction in search space for task-specific architectures is achieved by building on the optimized architectures for similar tasks instead of doing a full search without using this side information. Experimental results demonstrate the efficacy of the proposed approach and its improvements over the state-of-the-art methods.