Evolution under Length Constraints for CNN Architecture design
This reduces resource barriers for researchers and engineers in automated neural architecture search, though it is incremental as it builds on existing evolutionary methods.
The paper tackles the high computational cost of evolutionary CNN architecture design by proposing an evolution method with length constraints, achieving a 5.12% error rate on CIFAR-10 with only 4.6 GPU days, a reduction of 22 GPU days compared to prior work.
In recent years, the CNN architectures designed by evolution algorithms have proven to be competitive with handcrafted architectures designed by experts. However, these algorithms need a lot of computational power, which is beyond the capabilities of most researchers and engineers. To overcome this problem, we propose an evolution architecture under length constraints. It consists of two algorithms: a search length strategy to find an optimal space and a search architecture strategy based on genetic algorithm to find the best individual in the optimal space. Our algorithms reduce drastically resource cost and also keep good performance. On the Cifar-10 dataset, our framework presents outstanding performance with an error rate of 5.12% and only 4.6 GPU a day to converge to the optimal individual -22 GPU a day less than the lowest cost automatic evolutionary algorithm in the peer competition.