Zhichao Lu and colleagues accepting the Best Paper Award at GECCO 2019
Congratulations to BEACONites Zhichao Lu, Ian Whalen, Vishnu Boddeti, Yashesh Dhebar, Kalyanmoy Deb, Erik Goodman, and Wolfgang Banzhaf! Their paper “NSGA-Net: Neural Architecture Search using Multi-Objective Genetic Algorithm” won the Best Paper Award in the Evolutionary Machine Learning track at GECCO 2019 in Prague.
There were in total 64 papers submitted to the Evolutionary Machine Learning (EML) track, only 16 of which were accepted as full papers. Two papers were nominated for Best Paper Award. Zhichao Lu and colleagues won the award based on the on-site voting from the conference attendees.
Here is the abstract of the paper, which can be accessed from arXiv: https://arxiv.org/abs/1810.03522
This paper introduces NSGA-Net – an evolutionary approach for neural architecture search (NAS). NSGA-Net is designed with three goals in mind: (1) a procedure considering multiple and conflicting objectives, (2) an efficient procedure balancing exploration and exploitation of the space of potential neural network architectures, and (3) a procedure finding a diverse set of trade-off network architectures achieved in a single run. NSGA-Net is a population-based search algorithm that explores a space of potential neural network architectures in three steps, namely, a population initialization step that is based on prior-knowledge from hand-crafted architectures, an exploration step comprising crossover and mutation of architectures, and finally an exploitation step that utilizes the hidden useful knowledge stored in the entire history of evaluated neural architectures in the form of a Bayesian Network. Experimental results suggest that combining the dual objectives of minimizing an error metric and computational complexity, as measured by FLOPs, allows NSGA-Net to find competitive neural architectures. Moreover, NSGA-Net achieves a comparable error rate on the CIFAR-10 dataset when compared to other state-of-the-art NAS methods while using orders of magnitude less computational resources. These results are encouraging and show the promise to further use of EC methods in various deep-learning paradigms.
The source code for the paper can be accessed from GitHub: https://github.com/ianwhale/nsga-net
