BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search

TLDR

Neural architecture search has attracted many methods, and Bayesian optimization with a neural predictor has emerged as a promising strategy, yet prior studies focus on the full algorithm, obscuring which components drive performance. The authors aim to dissect the BO‑plus‑neural‑predictor framework by isolating its five key components—architecture encoding, neural predictor, uncertainty calibration, acquisition function, and acquisition optimization—to clarify their individual contributions to NAS performance. They evaluate multiple strategies for each component and introduce a novel path‑based encoding that theoretically and empirically outperforms existing encodings. Combining the best component choices yields BANANAS, a state‑of‑the‑art NAS algorithm whose code is publicly available.

Abstract

Over the past half-decade, many methods have been considered for neural architecture search (NAS). Bayesian optimization (BO), which has long had success in hyperparameter optimization, has recently emerged as a very promising strategy for NAS when it is coupled with a neural predictor. Recent work has proposed different instantiations of this framework, for example, using Bayesian neural networks or graph convolutional networks as the predictive model within BO. However, the analyses in these papers often focus on the full-fledged NAS algorithm, so it is difficult to tell which individual components of the framework lead to the best performance. In this work, we give a thorough analysis of the "BO + neural predictor framework" by identifying five main components: the architecture encoding, neural predictor, uncertainty calibration method, acquisition function, and acquisition function optimization. We test several different methods for each component and also develop a novel path-based encoding scheme for neural architectures, which we show theoretically and empirically scales better than other encodings. Using all of our analyses, we develop a final algorithm called BANANAS, which achieves state-of-the-art performance on NAS search spaces. We adhere to the NAS research checklist (Lindauer and Hutter 2019) to facilitate best practices, and our code is available at https://github.com/naszilla/naszilla.