Lilian Weng • 8/6/2020

Neural Architecture Search

This article provides a technical introduction to Neural Architecture Search (NAS), a field focused on automating the design of neural network architectures. It explains the three key components of a NAS system: the search space of operations, the search algorithm for sampling candidates, and the evaluation strategy for measuring performance. The content references seminal papers and discusses common representations like sequential layer-wise operations, positioning it as a detailed primer on this machine learning methodology.

0 comments

#Machine Learning #Neural Networks #Deep Learning