Noisy Heuristics NAS: A Network Morphism based Neural Architecture Search using Heuristics

TL;DR

This paper introduces Noisy Heuristics NAS, a neural architecture search method that uses biologically inspired heuristics to adapt network capacity dynamically, achieving performance comparable to ResNet-18 on standard datasets.

Contribution

It presents a novel NAS approach that employs learned heuristics for neuron addition and pruning, enabling online capacity adjustment based on biological neuronal dynamics.

Findings

Achieves competitive accuracy on MNIST, CIFAR-10, CIFAR-100.

Effectively adjusts network capacity during training.

Performs comparably to ResNet-18 with similar parameters.

Abstract

Network Morphism based Neural Architecture Search (NAS) is one of the most efficient methods, however, knowing where and when to add new neurons or remove dis-functional ones is generally left to black-box Reinforcement Learning models. In this paper, we present a new Network Morphism based NAS called Noisy Heuristics NAS which uses heuristics learned from manually developing neural network models and inspired by biological neuronal dynamics. Firstly, we add new neurons randomly and prune away some to select only the best fitting neurons. Secondly, we control the number of layers in the network using the relationship of hidden units to the number of input-output connections. Our method can increase or decrease the capacity or non-linearity of models online which is specified with a few meta-parameters by the user. Our method generalizes both on toy datasets and on real-world data sets such as MNIST, CIFAR-10, and CIFAR-100. The performance is comparable to the hand-engineered architecture ResNet-18 with the similar parameters.