Neural architecture search using progressive evolution

Nilotpal Sinha, Kuan Wen Chen

研究成果: Conference contribution同行評審

3 引文 斯高帕斯(Scopus)

摘要

Vanilla neural architecture search using evolutionary algorithms (EA) involves evaluating each architecture by training it from scratch, which is extremely time-consuming. This can be reduced by using a supernet to estimate the fitness of every architecture in the search space due to its weight sharing nature. However, the estimated fitness is very noisy due to the co-adaptation of the operations in the supernet. In this work, we propose a method called pEvoNAS wherein the whole neural architecture search space is progressively reduced to smaller search space regions with good architectures. This is achieved by using a trained supernet for architecture evaluation during the architecture search using genetic algorithm to find search space regions with good architectures. Upon reaching the final reduced search space, the supernet is then used to search for the best architecture in that search space using evolution. The search is also enhanced by using weight inheritance wherein the supernet for the smaller search space inherits its weights from previous trained supernet for the bigger search space. Experimentally, pEvoNAS gives better results on CIFAR-10 and CIFAR-100 while using significantly less computational resources as compared to previous EA-based methods. The code for our paper can be found here.

原文English
主出版物標題GECCO 2022 - Proceedings of the 2022 Genetic and Evolutionary Computation Conference
發行者Association for Computing Machinery, Inc
頁面1093-1101
頁數9
ISBN(電子)9781450392372
DOIs
出版狀態Published - 8 7月 2022
事件2022 Genetic and Evolutionary Computation Conference, GECCO 2022 - Virtual, Online, United States
持續時間: 9 7月 202213 7月 2022

出版系列

名字GECCO 2022 - Proceedings of the 2022 Genetic and Evolutionary Computation Conference

Conference

Conference2022 Genetic and Evolutionary Computation Conference, GECCO 2022
國家/地區United States
城市Virtual, Online
期間9/07/2213/07/22

指紋

深入研究「Neural architecture search using progressive evolution」主題。共同形成了獨特的指紋。

引用此